TH-1171 07610212

STATE SPACE BASED LOAD FREQUENCY CONTROL OF
MULTI-AREA POWER SYSTEMS
KRISHNA PAL SINGH PARMAR

STATE SPACE BASED LOAD FREQUENCY
CONTROL OF MULTI-AREA POWER
SYSTEMS
Thesis Submitted
in Partial Fulfilment of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
By
Krishna Pal Singh Parmar
Department of Electronics and Electrical Engineering

Indian Institute of Technology Guwahati
Guwahati - 781 039, INDIA.
October, 2013
TH-1171_07610212
Certificate
This is to certify that the thesis entitled “STATE SPACE BASED LOAD FREQUENCY
CONTROL OF MULTI-AREA POWER SYSTEMS”, submitted by Krishna Pal Singh
Parmar (07610212), a research scholar in the Department of Electronics & Electrical Engi-
neering, Indian Institute of Technology, Guwahati, for the award of the degree of Doctor of
Philosophy, is a record of an original research work carried out by him under my supervision
and guidance. The thesis has fulfilled all requirements as per the regulations of the Institute and
in my opinion has reached the standard needed for submission. The results embodied in this
thesis have not been submitted to any other University or Institute for the award of any degree
or diploma.
Dated: Prof. Somanath Majhi

Guwahati. Dept. of Electronics & Electrical Engg.
Indian Institute of Technology Guwahati
Guwahati - 781039, Assam, India.
TH-1171_07610212
Acknowledgements
I would like to express my sincere gratitude to my supervisor, Prof. Somanath Majhi, for his
excellent guidance and support throughout the course for this work. My heartfelt thanks to him
for the unlimited support and patience shown to me. I would particularly like to thank for all his
help in patiently and carefully correcting all my manuscripts and thesis. I would like to thank
my doctoral committee members Prof. P. K. Bora, Prof. C. Mahanta and Dr. H. B. Nemade for
sparing their precious time to evaluate the progress of my work. Their suggestions have been
valuable. I am thankful to Prof. D. P. Kothari, Director(Research), MVSR Engineering College,
Hyderabad, India and former Director i/c, IIT Delhi for guiding me on power system issues and
sparing his valuable time to solve my difficulties all the way. I would also like to thank other
faculty members for their kind help during my academic studies.

My special thanks to Mr. Sanjib, Mr. Sidananda, Mr. Goswami and all the members of the
C&I Laboratory for providing various resources useful in research work. I had a great time with
my many friends at IITG, including (but not limited to) Mr. Sanjoy, Mr. Dola, Mr. Govind, Mr.
Bajrangbali and Mr Mridul. I thank them for their support and encouragement.
Among those deserving of a special mention is my wife Atri Parmar, who had extended
her wholeheartedly support to bring this work into existence. She has been able to take care of
my daughter and son extremely well. I thank my daughter Vaidehi and son Vansh for keeping
me energetic and reducing my worries. Again, I am grateful to my parents, mother in-law Ms
Shyama Baghel and elder brother Mr. R. B. Singh Parmar, whose love and encouragement
made this research possible.
I am grateful to Sh. Subodh Garg, Director General, Sh. J. S. S. Rao, Principal Director
(CP), Sh. S. K. Choudhary, Principal Director(MS), Ms Manju Mam, Director (MS) and col-
leagues of NPTI for guiding and helping me on various technical aspects. My special thanks
goes to Sh. Manas Ray, former Director, NPTI, Guwahati who helped me to initiate this re-
search work. Finally, I would like to thank the Almighty God for bestowing me this opportunity
and showering his blessings on me to come out successful against all odds.
(Krishna Pal Singh Parmar)

TH-1171_07610212
Abstract
The main aim of this thesis is to present a state space based load frequency control (LFC)
system in conventional and restructured power system environment emphasizing on multi-
source power generation (MSPG). An output feedback controller (OFC) is presented with a
pragmatic point of view. As full state feedback controllers (FSFCs) require the transfer of in-
formation from all parts of the system to a central control facility for processing, which for large
scale interconnected power system could prove to be prohibitive, more practical and alternative
forms of feedback controllers such as OFCs have been the subject of extensive investigation in
the various control engineering applications. As these controllers use only a subset of the state
vector for feedback purposes, they are simpler, more practical and easy to implement than the
FSFCs. A simple algorithm is presented to optimize the OFC gains. All the power system mod-
els presented for LFC study have been simulated using MATLAB Simulation tool and dynamic
responses, thereof, are obtained at optimum controller gain settings. The dynamic responses
obtained with OFC have been compared with that of FSFC and in most of the cases improved
results are obtained.

Most of the researchers worked on LFC of power systems considering thermal unit with
non-reheat turbines. In this thesis, the LFC is extended for the conventional power systems
with the combinations of non-reheat turbines, reheat turbines and hydro turbine. The control
area having MSPG represented by an equivalent of thermal or hydro unit dynamics only may
not result in a realistic design of LFC control. Therefore, new power system models with
MSPG are presented in this thesis for LFC study in both conventional and restructured power
system environment. The LFC system is further improved by considering AC-DC tie lines and
thyristor controlled phase shifter (TCPS). Modified LFC of an interconnected power system
with MSPG is proposed in restructured power system environment considering all the possible
contracts between generation companies (GENECOs) and distribution companies(DISCOs).
The LFC of hydro power plants operational in KHOZESTAN, IRAN has also been studied.
TH-1171_07610212
The proposed controller performs well on this hydro plant and improves the frequency deviation
responses remarkably. The proposed controller shows its capability and flexibility by providing
the desirable dynamic responses to all the power system models studied. To examine the robust
performance of the OFC, the system parameters and operating load conditions are varied by
±25% from their nominal values to obtain the dynamic responses. The effects of generation rate
constraint (GRC) and variation in step load perturbations (SLPs) are also examined. Dynamic
responses are obtained by varying the regulation parameter and the most appropriate value is
suggested for the power systems with MSPG.
Many research work on LFC systems have been reported in recent years. Still there is much
room for further improvement and extensions of such schemes in new power system environ-
ment. In this thesis, LFC systems of several new models have been presented and improved
dynamic responses are obtained.
TH-1171_07610212
2
C ONTENTS
List of Figures iv
List of Tables vii
Nomenclature ix
Mathematical Notations x
List of Publications xiii
1 Introduction 1
1.1 Load frequency control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Research Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 Significant contributions of the thesis . . . . . . . . . . . . . . . . . . . . . . . 10

1.5 Organization of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2 Mathematical modeling and LFC of power systems 13

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Fundamental LFC loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Frequency response modeling . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 LFC modeling of the MAIPS . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.5 LFC modeling of MAIPS with MSPG . . . . . . . . . . . . . . . . . . . . . . 26
2.6 Controller Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.6.1 Algorithm for OFC . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.6.2 Selection of Q̃ and R̃ matrices . . . . . . . . . . . . . . . . . . . . . . 32
TH-1171_07610212
i
Contents
2.7 Important steps to solve LFC problem . . . . . . . . . . . . . . . . . . . . . . 32
2.8 State-space model of the power systems . . . . . . . . . . . . . . . . . . . . . 33

2.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 LFC of conventional power systems 35

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 TAIPS considering non-reheat turbines . . . . . . . . . . . . . . . . . . . . . . 35
3.2.1 State-space Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.2 Simulation results and discussion . . . . . . . . . . . . . . . . . . . . 38

3.3 TAIPS considering reheat and non-reheat turbines . . . . . . . . . . . . . . . . 40
3.4 Hydro-thermal interconnected power system . . . . . . . . . . . . . . . . . . . 44

3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4 LFC of power systems with multi-source power generation 51

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2 Single area power system with MSPG . . . . . . . . . . . . . . . . . . . . . . 52
4.2.2 Effect of generation rate constraint . . . . . . . . . . . . . . . . . . . . 55

4.2.3 Effect of governor speed regulation parameter (R) . . . . . . . . . . . . 56
4.2.4 Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.3 Two-area interconnected power system with MSPG . . . . . . . . . . . . . . . 60

4.3.2 Effect of GRC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.3.3 Selection of governor speed regulation parameter(R) . . . . . . . . . . 67
4.3.4 Effect of variation in load disturbance . . . . . . . . . . . . . . . . . . 67
4.3.5 Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

4.4 LFC of hydro power plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
TH-1171_07610212
ii
Contents
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5 LFC of power systems considering AC-DC tie lines and TCPS 78
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.2 Simulation of the power system with parallel AC-DC tie lines . . . . . . . . . . 80
5.3 Incremental tie line power flow considering TCPS . . . . . . . . . . . . . . . 84
5.3.1 TCPS control strategy . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.3.2 Simulation of the power system model considering TCPS . . . . . . . 89

5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6 LFC in restructured power system environment 98

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.2 LFC in restructured power system environment . . . . . . . . . . . . . . . . . 99
6.3 Simulation and results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

6.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
7 Conclusions and Future work 115

7.1 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
7.2 Suggestions for further work . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
A Matrices of power systems 119

A.1 Matrices of TAIPS with MSPG . . . . . . . . . . . . . . . . . . . . . . . . . . 119
A.2 Matrices of TAIPS with MSPG considering AC-DC tie lines . . . . . . . . . . 122
A.3 Matrices of TAIPS with MSPG considering TCPS . . . . . . . . . . . . . . . . 125
A.4 Matrices of TAIPS with MSPG in restructured power system . . . . . . . . . . 128
References 133
TH-1171_07610212
iii
L IST OF F IGURES
2.1 Schematic block diagram of a synchronous generator with basic frequency con-
trol loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2 Linear model of the primary LFC loop . . . . . . . . . . . . . . . . . . . . . . 18

2.3 Linear model with supplementary LFC loop . . . . . . . . . . . . . . . . . . . 21
2.4 N-control areas power system . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.5 Block diagram representation for tie line power change of control area-i in an
interconnected power system with N-control areas . . . . . . . . . . . . . . . . 24

2.6 Control area-i of an interconnected power system with N number of control areas 25
2.7 Control area-i of MAIPS with MSPG and N-number of control areas . . . . . . 26
3.1 TAIPS with non-reheat turbines . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2 Dynamic responses of TAIPS with non-reheat turbines . . . . . . . . . . . . . 41

3.3 TAIPS considering reheat and non-reheat turbines . . . . . . . . . . . . . . . . 42
3.4 Dynamic responses of TAIPS with reheat and non-reheat turbine . . . . . . . . 45
3.5 Hydro-thermal interconnected power system . . . . . . . . . . . . . . . . . . . 47
3.6 Dynamic response of hydro-thermal interconnected power system . . . . . . . 49
4.1 Block diagram of the single area Power System comprising Reheat-Thermal,
Hydro and Gas generating units . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2 Frequency deviation response to 1% SLP in the area . . . . . . . . . . . . . . 56
4.3 Power generation deviation responses to 1% SLP in the area . . . . . . . . . . 57
4.4 Frequency deviation response to 1% SLP in the area considering GRC . . . . . 58

4.5 Block diagram of a non-linear turbine(GRC) . . . . . . . . . . . . . . . . . . . 58
TH-1171_07610212
iv
List of Figures
4.6 Frequency deviation response to 1% SLP in the area with variation in R from
1% to 8% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.7 Frequency deviation response with variation in load conditions . . . . . . . . . 60
4.8 Frequency deviation response with variation in speed governor time constant . 60
4.9 TAIPS comprising reheat-thermal, hydro and gas generating units in each con-
trol area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.10 Comparison of frequency and tie line power deviation responses for 1% SLP . 64
4.11 Frequency and tie line power deviation responses with low, medium and high
head hydro turbines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.12 Dynamic responses considering the GRC . . . . . . . . . . . . . . . . . . . . 66

4.13 Effect of the R on dynamic responses . . . . . . . . . . . . . . . . . . . . . . 68
4.14 Frequency and tie line power deviation responses for SLPs in area-1, varying
from 1% to 5% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.15 Effect of variation in steam turbine speed governor time constant(TSG ) on dy-
namic responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.16 Effect of variation in steam turbine time constant(TT ) on dynamic responses . . 70
4.17 Effect of variation in steam turbine reheat time constant(TR ) on dynamic re-
sponses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.18 Effect of variation in hydro turbine speed governor main servo time constant(TGH )
on dynamic responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.19 Effect of variation in hydro turbine speed governor transient droop time constant(TRH )
on dynamic responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.20 Effect of variation in gas turbine compressor discharge volume time constant
(TCD ) on dynamic responses . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.21 Effect of variation in nominal starting time of water in penstock (TW ) on dy-
namic responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.22 Effect of variation in nominal area load on dynamic responses . . . . . . . . . 72
4.23 Model of hydro power plants operational in KHOZESTAN (Iran), [1] . . . . . 74
TH-1171_07610212
v
List of Figures
4.24 (a)Area load disturbance w1 , first situation (b) Disturbance due to neighboring
interconnected control areas w2 , second situation [1] . . . . . . . . . . . . . . 74

4.25 Dynamic responses of hydro power plants operational in KHOZESTAN (Iran) . 76
5.1 TAIPS with AC-DC parallel tie lines . . . . . . . . . . . . . . . . . . . . . . . 80

5.2 Block diagram of a TAIPS with MSPG considering AC-DC parallel tie lines . . 81
5.3 Frequency deviation response of control area-1 with AC tie line and AC-DC
parallel tie line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.4 Frequency deviation response of control area-2 with AC tie line and AC-DC
parallel tie line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.5 Tie line power deviation response with AC tie line and AC-DC parallel tie line 85
5.6 Frequency deviation and tie line power responses for SLP varying from 1% to 4% 86
5.7 TAIPS with MSPG considering TCPS . . . . . . . . . . . . . . . . . . . . . . 87
5.8 Schematic of TAIPS considering TCPS in series with tie line . . . . . . . . . . 87
5.9 Frequency deviation response of control area-1 . . . . . . . . . . . . . . . . . 93
5.10 Frequency deviation response of control area-2 . . . . . . . . . . . . . . . . . 93

5.11 Tie line power deviation response . . . . . . . . . . . . . . . . . . . . . . . . 94
5.12 TCPS phase shift deviation response . . . . . . . . . . . . . . . . . . . . . . . 94
5.13 Frequency deviation and tie line power responses for SLP varying from 1% to 4% 95
6.1 A TAIPS with MSPG in restructured power system environment . . . . . . . . 101

6.2 Frequency and tie line power deviation responses for case-1 . . . . . . . . . . . 106
6.3 Generator power output response for case-1 . . . . . . . . . . . . . . . . . . . 107

TH-1171_07610212
vi
L IST OF TABLES
3.1 Simulation parameters of TAIPS considering non-reheat turbines . . . . . . . . 39

3.2 Dynamic response comparison in terms of peak overshoot (OS) . . . . . . . . . 40
3.3 Dynamic response comparison in terms of ST . . . . . . . . . . . . . . . . . . 40

3.4 Simulation parameters of TAIPS considering reheat and non-reheat turbines . . 43
3.7 Simulation parameters of hydro-thermal interconnected power system . . . . . 46

4.1 Simulation parameters of SAPS with MSPG . . . . . . . . . . . . . . . . . . . 55
4.2 Dynamic response comparison of a SAPS . . . . . . . . . . . . . . . . . . . . 56

4.3 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4 Simulation parameters of TAIPS with MSPG . . . . . . . . . . . . . . . . . . 61
5.1 Simulation Parameters of TAIPS with MSPG considering parallel AC-DC tie
lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.2 Power Generation Scheduling to match the nominal load of the individual con-
trol area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.3 Dynamic response comparison . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.4 Simulation Parameters of TAIPS model with MSPG considering TCPS . . . . . 90
5.5 Power Generation Scheduling to match the nominal load of the individual con-
trol area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.6 Dynamic response comparison in terms of peak OS . . . . . . . . . . . . . . . 92
TH-1171_07610212
vii
List of Tables
6.1 Simulation parameters of TAIPS with MSPG in restructured power environment 104
TH-1171_07610212
viii
N OMENCLATURE
ACE Area control error
AGC Automatic generation control

DISCO Distribution company
FSFC Full state feedback controller
GENCO Generation company
ISE Integral square error

ITAE Integral of time weighted absolute error
LFC Load frequency control
MAIPS Multi-area interconnected power system
MIMO Multi-input multi-output
MISO Multi-input single output

MSPG Multi-source power generation
OFC Output feedback controller
OS Overshoot
PI Proportional integral controller

PID Proportional integral and derivative controller
SAPS Single area power system
SLP Step load perturbation
ST Settling time
TAIPS Two area interconnected power system
TCPS Thyristor controlled phase shifter
TRANSCO Transmission company
TH-1171_07610212
ix
M ATHEMATICAL NOTATIONS
ACEi Area control error
Prti Rated capacity of control area, MW

PLi Nominal load of control area, MW
f Nominal system frequency, Hz
∆ fi Incremental change in frequency of control area, Hz
Yg (s) Governor dynamics

Yt (s) Turbine dynamics
Yp (s) Power system(rotating mass and load) dynamics
∆Ptie,12 Incremental change in actual tie line power flow from control area-1 to 2, pu MW
∆Ptie,i Total tie line power change between control area-i and all other control areas, pu MW
∆PDi Total incremental change in local load of control area, pu MW

∆Pmi Incremental generation change in turbine power output, pu MW
∆Pmki Incremental generation change in turbine power output of kth GENCO of area, pu MW
∆Pgi Incremental change in governor output command, pu MW
∆Pvi Incremental change in governor valve position, pu MW

∆Pci Incremental change in speed changer position, pu MW
∆Pcki Incremental change in speed changer position of kth GENCO of area, pu MW
Di Equivalent system damping coefficient of control area, pu MW/Hz
Hi Equivalent inertia constant of control area, MW-s/MVA

Bi Frequency bias parameter of control area, pu MW/Hz
Ki (s) Controller dynamics
K̃ Output feedback controller gain matrix
TH-1171_07610212
x
Mathematical Notations
K̄ Full state feedback controller gain matrix
a12 control area capacity ratio

TSGi Steam turbine speed governor time constant, s
TTi Steam turbine time constant, s
TPSi Power system time constant, s
Ri Governor speed regulation parameter of control area, Hz/pu MW
Rki Governor speed regulation parameter of kth GENCO of area, Hz/pu MW

KPSi Power system gain, Hz/pu MW
KRi Steam turbine reheat constant
TRi Steam turbine reheat time constant, s
TWi Nominal starting time of water in penstock, s

TRSi Hydro turbine speed governor reset time, s
TRHi Hydro turbine speed governor transient droop time constant, s
TGHi Hydro turbine speed governor main servo time constant, s
XGi Lead time constant of Gas turbine speed governor, s

YGi Lag time constant of Gas turbine speed governor, s
cgi Gas turbine valve positioner
bgi Gas turbine constant of valve positioner, s
TFi Gas turbine fuel time constant, s
TCRi Gas turbine combustion reaction time delay, s

TCDi Gas turbine compressor discharge volume -time constant, s
αki Generation contribution factor of kth GENCO of area
ep fk Economic participation factor of kth generating unit
cp fkl Contract participation factor between kth GENCO and l th DISCO

T12 Tie line power coefficient between area-1 and area-2
KΦ Gain constant of TCPS
TΦ Time constant of TCPS, s
KDCi Gain constant of HVDC link
TH-1171_07610212
xi
Mathematical Notations
TDCi Time constant of HVDC link, s
Subscripts
i ith control area
k kth GENCO
l l th DISCO
ki kth GENCO of ith control area
TH-1171_07610212
xii
L IST OF P UBLICATIONS
Journal Publications
1. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Optimal Load Frequency Control of an In-
terconnected Power System, MIT International Journal of Electrical and Instrumentation
Engineering, pp 1-5, Vol-1, No.1, 2011
2. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Improvement of Dynamic Performance of

LFC of the Two Area Power System: an Analysis using MATLAB, International Journal
of Computer Applications, Vol 40, No. 10, pp 28-32, 2012
3. K. P. Singh Parmar, S. Majhi and D. P. Kothari, LFC of an Interconnected Power System

with Thyristor Controlled Phase Shifter in the Tie Line, International Journal of Computer
Applications, Vol 41, No. 9, pp 27-30, 2012
4. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Load frequency control of a realistic

power system with multi-source power generation, International Journal of Electrical
Power and Energy Systems (Elsevier), Vol.42, pp 426-33, 2012
5. K. P. Singh Parmar, S. Majhi and D. P. Kothari, LFC of an Interconnected Power System

with Multi-source Power Generation in Deregulated Power Environment, International
Journal of Electrical Power and Energy Systems (Elsevier), manuscript no. IJEPES-D-
12-00804R1 (Revision submitted)
Conference Publications
1. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Multi-Area Load Frequency Control

in a Power System Using Optimal Output Feedback Method, IEEE Conf. proceedings,
PEDES, New Delhi, India, 2010.
TH-1171_07610212
xiii
List of Publications
2. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Automatic Generation Control of an Inter-
connected Hydrothermal Power System, IEEE Conf. proceedings, INDICON, Kolkata,

India, 2010.
3. K. P. Singh Parmar, S. Majhi and D. P. Kothari, Automatic Generation Control Strategies:

A state of art survey, International Conference on Advances in Electrical and Electronics
Engineering (ICAEEE) Proceedings, Moradabad, India, 2011.
TH-1171_07610212
xiv
C HAPTER 1
I NTRODUCTION
1.1 Load frequency control
The automatic generation control(AGC) or LFC problem in power systems has a long history
and has been one of the most important topics of research and study of interconnected power
systems. In a power system, LFC as an ancillary service plays an important and fundamental
role to maintain the power system reliability at an appropriate level. It has gained the im-
portance with the change of power system structure and the growth of size and complexity
of interconnected systems [1–9]. The successful operation of interconnected power systems
requires the matching of total power generation with total load demand and associated power
system losses [2–4]. As the demand deviates from its nominal value with an unpredictable small
amount, the operating point of power system changes, and hence, system may experience devi-
ations in nominal system frequency and scheduled tie line power exchanges, which may yield
undesirable effects [2, 7, 9, 10]. Maintaining system frequency and tie line power interchanges
with neighboring control areas at the scheduled values are the two main objectives of a power
system LFC [4, 7, 10–14]. These objectives are met by measuring a control error signal, called
the area control error (ACE), which represents the real power imbalance between generation
and load [2, 6, 14, 15].
Depending on the type of generating units, and constraints on their range and rate of re-
sponse to the LFC signals, the actual response time varies from a few seconds to several sec-
onds [14, 16, 17]. In LFC practice, rapidly varying components of system signals are almost
unobservable due to filters involved in the process. That is why further reduction in the re-
TH-1171_07610212
1
Chapter 1 Introduction
sponse time of LFC is neither possible nor desired. In practice, the design and performance of
an LFC system depend on how generation units respond to control signal. Such control schemes
are useful as they are able to maintain a sufficient level of reserved control range and a sufficient
level of control rate [17].
Global analysis of the power system markets shows that LFC is one of the most profitable
ancillary services of the interconnected power systems. This service is related to the short-term
balance of real power and frequency of the power systems and acquires a principal role to enable
power exchanges and to provide better conditions for electricity trading [5, 6, 11, 18–22]. The
LFC modeling and control scheme have been explained in detail in the Chapter 2.
1.2 Research Background
LFC plays an important role in modern energy management systems. The LFC problem has
been extensively studied during the last few decades. Most of the work focused on the LFC of
conventional single area or two area power systems. The first attempt in the area of LFC has
been to control the frequency of a power system via the flywheel governor of the synchronous
machine. This technique was subsequently found to be insufficient, and a supplementary control
was included to the governor with the help of a signal directly proportional to the frequency
deviation plus its integral. This scheme constitutes the classical approach to the LFC of power
systems. Very early works in this important area of LFC have been by Cohn et al. [23–27].
These works were based on tie line bias control strategy. Quazza [28] illustrated non-interactive
control considering (i) non-interaction between frequency and tie line powers controls and (ii)
each control area taking care of its own load variations. The revolutionary optimal control
concept for LFC regulator designs of interconnected power systems was presented by Elgerd
[8].
A technique based on coordinated system wide correction of time error and inadvertent
interchange was discussed in an LFC study by Cohn [29]. Supplementary controllers were de-
signed to regulate the ACEs to zero effectively. The research work also contributed the design
TH-1171_07610212
2
of LFC based on various control techniques. The power systems are usually large scale systems
with complex nonlinear dynamics. However, most of the research work reported so far has
been performed by considering linearized models of interconnected power systems. The LFC
response for small load perturbations using linearized models of various power system compo-
nents has been widely accepted [1–10]. The effect of GRC was included in the LFC studies,
considering both continuous and discrete power system models [30–34]. Lot of research work
has been presented on the modeling of various energy source dynamics for LFC studies of
power systems [2, 5–7, 10, 35–38].
The LFC regulator design methods using modern optimal control theory enable the power
engineers to design an optimal control system with respect to given performance criterion.
Fosha and Elgerd [39] were the first to present their work on optimal LFC regulator design
using this concept. A two-area interconnected power system (TAIPS)consisting of two identi-
cal power plants of non-reheat thermal turbines was taken for investigations. A new formulation
for optimal LFC strategy has been proposed in [40]. The feasibility of an optimal LFC scheme
requires the availability of all state variables for feedback. Next, the problem is to reconstruct
the unavailable states from the available outputs and controls using an observer.
Considering state reconstruction, many significant contributions have been made [41–46].
Bohn and Miniesy [41] have studied the LFC of a TAIPS by making the use of differential ap-
proximation and a Luenberger observer. Exploiting the fact that the nonlinearity of the power
system model, namely, the tie line power flow, is measurable, the observer has been designed
to give zero asymptotic error, even for the nonlinear model. LFC schemes based on an optimal
observer, using a nonlinear transformation [42] and reduced-order system models with a local
observer [43] have been reported in the literature. A simple generating unit model oriented to-
wards LFC and the method for its transfer function identification based on a two-stage procedure
is presented in [46]. Due to practical difficulties in the implementation of regulators based on
feedback of all state variables, suboptimal LFC regulator designs were considered [47–49]. A
suboptimal and near-optimal LFC concept using modern control theory is proposed by Moorthy
and Aggarwal [47]. The LFC regulator design using Lyapunov’s second method and utilizing
TH-1171_07610212
3
minimum settling time theory has been proposed by Shirai [50]. The design of decentralized
LFC based on structured singular values is presented in [51].

Various LFC schemes based on two-level [52] and multi-level [53–55] control concepts have
been proposed. A two-level suboptimal controller has been suggested by Miniesy and Bohn
[52]. However, this approach does not ensure zero steady state error, and hence, a multilevel
finite time optimal controller design ensuring zero steady state error has been reported in [53].
The advantage of hierarchical structure is reflected in the fact that even if one of the control
levels fails, the system remains in operation.
Keeping in view the accuracy and reliability of digital controllers, researchers have focussed
on proposing digital LFC control schemes [56–63]. Ross [56] was probably the first to present
a comprehensive direct digital LFC regulator for the power systems. Later, a digital LFC in-
corporating dynamic control criteria for performance evaluation of digital control system based
on field test results was initiated by Ross and Green in [57]. Bohn and Miniesy [41] have
analyzed the effect of the sampling period on the system’s dynamic response. Kothari and
coworkers [62, 64] have studied the LFC in discrete mode. The investigations were carried out
with more feasible modeling of LFC strategy, i.e., considering that the system is operating in
continuous mode and the controller is operating in discrete mode [64]. In [62], discrete mode
LFC of an interconnected power system with reheat turbines considering a new ACE is de-
scribed. The new ACE is derived from tie line power deviation, frequency deviation, time error,
and inadvertent interchanges.
A robust controller design based on the Riccati equation approach has been proposed for
the power systems [65, 66]. Later, based on a combination of the robust control approach
and an adaptive control technique, a design procedure of a new robust adaptive controller was
proposed for power system LFC with system parametric uncertainties [65]. The other research
contributions on decentralized robust LFC based on the Riccati equation approach have been
recorded in [67]. The design of decentralized robust LFC applying structured singular values
is proposed by Yang et al. [68]. It has been shown that when the frequency response based
diagonal dominance cannot be achieved, the structured singular values can be applied to design
TH-1171_07610212
4
decentralized LFC to achieve the desired system dynamic performance [68].
Apart from various LFC schemes, adaptive control has been a topic of research for a long
time. The task of adaptive control is to make the process under control less sensitive to changes
in process parameters and process dynamics. A number of articles have been appeared on
adaptive LFC schemes [69–74]. In 1966, Ross [69] described control criteria in LFC and the
related practical difficulties encountered in trying to achieve these criteria. The implementation
and analysis of an adaptive LFC system on the Hungarian power system has been proposed
by Vajk et al. [71]. A multi-area adaptive LFC scheme for LFC of power systems [73] and a
reduced-order adaptive LFC for interconnected hydro-thermal power system [74] are reported
in the literature.
Many researchers have applied the artificial intelligence and evolutionary algorithms based
controllers for the LFC study. In recent years, the advent of modern intelligent methods, such
as Artificial neural networks (ANNs), fuzzy logic, Genetic algorithms(GAs), Particle swarm
optimization (PSO), PSO-hybrids and Bacterial Foraging(BF)-based optimization has solved
the LFC problems [4, 15, 75–81]. Application of the ANN technique based on robust control
methodologies for solution of the LFC problem in interconnected power system has appeared
in the literatures [75, 76]. Shayeghi and Shayanfar [75] have presented the concept of the H∞
robust control technique for training of radial biased function (RBF) neural networks for im-
provement of the performance of the proposed controller under various operating conditions. In
Ref. [76], the idea of µ -synthesis control techniques has been used for training an ANN-based
LFC controller too. Researches on the LFC problem shows the LFC analysis using the fuzzy
proportional plus integral (PI) controller [78, 79].
Application of the optimal control theory to power system has shown that an optimal load
frequency controller can improve the dynamic stability of a power system [4]. As state feedback
controllers require the transfer of information from all parts of the system to a central control
facility for processing, which for large scale interconnected power system could prove to be pro-
hibitive, alternative more practical forms of feedback controllers such as output feedback and
decentralized controllers have been the subject of extensive investigation. As these controllers
TH-1171_07610212
5
use only a subset of the state vector for feedback purposes, they are simpler, more practical and
easy to implement than the full state feedback controllers [2, 49]. Bettayeb et. al [82] proposed
the incorporation of time-weighted linear quadratic regulator state and output feedback control
design for power system dynamic stability analysis. Shahnazi et. al [83] presented an output
feedback control with disturbance rejection for a class of nonlinear multi-input multi- output
(MIMO) systems with unknown but bounded disturbances. An application of linear quadratic
Gaussian (LQG) based load frequency controller in a competitive electricity environment is
witness of the popularity of state feedback controls [11]. Tyagi and Srivastava [84] presents
the design of a decentralized AGC scheme where the controller has been designed by appropri-
ately assigning the eigen-structure of each isolated subsystem via state feedback, satisfying the
sufficient conditions for stability.
Most recently many researchers [1, 12, 13, 15, 80, 85–87] have studied the LFC problem of
hydro, thermal systems using proportional, integral and derivative (PID) controller, fuzzy con-
troller, decentralized controller and optimal multi input single output (MISO) PID controller
based on different algorithms and optimization techniques. The fuzzy PI controller is known to
give poor performance in system transient response. Chang et al. [88], obtained the optimum ad-
justment gains of the integral controller using GAs through performance indices integral square
error (ISE) and integral of time of absolute error (ITAE). The premature convergence of GA
degrades its efficiency and reduces the search capability. Nanda et al. [15] proposed the LFC
using Bacterial Foraging-Based Optimization. PSO is developed through simulation of bird
flocking in multi-dimensional space. Like GA, PSO is also less susceptible to getting trapped
on local optimum [15, 80]. Mojtaba et al. [85] proposed a robust multi-variable model based
predictive control (MPC) for the solution of LFC in a multi-area interconnected power system
(MAIPS). The proposed control scheme is designed to consider multi-variable nature of LFC,
system uncertainty and generation rate constraint, simultaneously. Alireza et al. [1] studied
the LFC of the hydro power system (operational in Iran) using optimal MISO PID controller.
Hasan et al. [86] presented the design of sub-optimal AGC regulators based on the constrained
feedback control strategy using the feedback of system states. Decentralized load frequency
TH-1171_07610212
6
controller is presented for the LFC of an interconnected thermal power system [13] which uses
large number of states for the controller feedback. Challa et al. [89] has presented the analysis
and design of controller for two area hydro-thermal-gas AGC system. They have shown that for
LFC study, optimal PI state feedback controller is more robust and performs better than conven-
tional genetic algorithm based PI controller. However, this optimal PI state feedback controller
uses all the states for feedback purpose which is practically difficult and results in the increased
complexity and cost of the controller. Rakhshani et al. [87] have applied reduced-order observer
control for two-area LFC system after deregulation. In fact, one of the main observed problems
in the control of AGC systems is the limitation to access and measurement of state variables in
the real world. So with a practical point of view, an optimal output feedback method, is used to
solve this problem [90–94]. In the output feedback method, only the measurable state variables
within each control area are required to use for feedback. All these controllers discussed have
relative advantages and disadvantages. Shayeghi et al. [95] concluded in the state of art survey
on LFC, that there are no rules as to when a particular technique is more suitable for the LFC
problem.
Apart from advances in control concepts, there have been many changes during the last
decade or more, such as deregulation of the power industry and use of super conducting mag-
netic energy storage, wind turbines and photovoltaic cells as other sources of electrical energy
to the system. In a power system, the instantaneous mismatch between supply and demand of
real power for sudden load changes can be reduced by the addition of active power sources
with fast response such as battery energy storage (BES), super conducting magnetic energy
storage (SMES), capacitive energy storage (CES) and Redox Flow Batteries devices [96–100].
Bhatti and Kothari [101] presented the Variable structure load-frequency control of isolated
wind-diesel-micro-hydro hybrid power systems.
Literature survey shows that mostly AC tie lines are used for the interconnection of multi-
area power systems and lesser attention is given to AC-DC parallel tie lines [2, 4]. HVDC
transmission has emerged due to its various techno-economical advantages. One of the major
applications of HVDC transmission is operating a DC link in parallel with an AC link intercon-
TH-1171_07610212
7
necting two control areas to get an improved system dynamic performance with greater stability
margins under small disturbances in the system [102,103]. Considerable research work has been
carried out on LFC of interconnected power systems connected via HVDC link in parallel with
AC link [4, 102–105].
The Flexible AC Transmission Systems (FACTS) devices provide more flexibility in power
system operation and control. TCPS is an effective FACTS device for the tie line power flow
control of an interconnected power system. The TCPS device is modeled and used in series
with tie lines to improve the dynamic performance of LFC of the interconnected power systems
by the many researchers [4, 106–109].
The classical LFC based on ACE is difficult to implement in a restructured power system
environment. In recent years, several control scenarios based on robust and optimal approaches
have been proposed for the AGC system in deregulated power systems. Some research is con-
tained in [5,6,11,18–22,84,110–113]. In the restructured power system environment, vertically
integrated system of conventional power system do not exist [20]. In a competitive electric-
ity market, generating companies (GENCOs), distribution companies (DISCOs), transmission

companies (TRANSCOs), and power system operator (PSO) [5, 6, 18, 19, 114] are all mar-
ket players. As there are so many GENCOs and DISCOs in the deregulated power system, a
DISCO has the freedom to have a contract with any GENCOs for the transaction of power. A
DISCO of one control area can make contract with a GENCO in another control area [19, 20].
For stable and secure operation of a power system, the PSO has to provide a number of ancil-
lary services. One of the ancillary services is the frequency regulation based on the concept
of the LFC. The crucial role of LFC system will continue in restructured power system envi-
ronment with some modifications accounting bilateral transactions and deregulation policy [5].
A detailed discussion on LFC issues in power system operation after deregulation is reported
in [20–22].
An LFC system required for Poolco-based transactions described in [20, 112] utilizes an
integral controller. A method to find optimal controller gains of this type of controller for
a two-area system is proposed in [20]. In [115], B. Tyagi et. al proposed a general model
TH-1171_07610212
8
for multi-area LFC suitable for a competitive electricity environment. LFC work in deregulated
power system is reported in [5,11,18–20,84] where they have considered either thermal or hydro
system in a control area. In new power system environment, a control area may have variety of
sources like hydro, thermal, gas, renewable etc., therefore representing a control area by thermal
or hydro system dynamics only may not result in a good design of LFC system [5, 6, 89, 90].
Recently some researchers studied the LFC of conventional power system considering hydro,
thermal and gas generating units in each control area [89, 90, 116, 117], however they did not
consider the LFC scheme in restructured power system environment.
1.3 Motivation
Global analysis of the power system markets shows that LFC is one of the most profitable
ancillary services of the interconnected power systems. LFC is very important topic of research
and many researchers studied this problems with various combinations of controllers and power
system models. All the controller methods presented in literature for LFC study have their own
advantages and disadvantages.
Application of the optimal control theory to power system has shown that an optimal LFC
can improve the dynamic stability of a power system. The LFC regulator design techniques
using modern optimal control theory enable the power engineers to design an optimal control
system with respect to given performance criterion. As state feedback controllers require the
transfer of information from all parts of the system to a central control facility for processing,
which for large scale inter connected power system could prove to be prohibitive. In the larger
systems all the states may not be available for measurement and require large number of sensors.
Keeping in view the above, researchers motivated to make application of more promising and
practical form of optimal OFC for the study of LFC as this controller uses only a subset of the
state vector for feedback purposes, this is simpler, more practical and easy to implement than
the full state feedback controller.
Apart from advances in control concepts, there have been many changes during the last
TH-1171_07610212
9
decade or more, such as deregulation of the power industry, use of AC-DC tie lines, FACTS
devices etc. Most of the researchers considered either thermal or hydro generating units in a
control area. In a real situation, control area may have variety of sources of generations such
as hydro, thermal, gas, nuclear, solar, wind etc. The control area having different sources of
power generation represented by an equivalent of thermal or hydro unit dynamics only may
not result in realistic design of LFC control. Keeping in view the changing power scenario,
combination of multi-source generators in a control area with their corresponding generation
contribution factors is more realistic for the study of LFC Further, this motivated to include
the more realistic combination of multi-source power generation (Thermal, hydro and gas) in a
control area. An attempt has been made to study the LFC of various new power system models
with MSPG including FACTS devices (TCPS), AC-DC tie lines and restructured power system
environment. To demonstrate the performance of controller over a wide range of variation in
parameters and load condition is also one of the key factor of motivation.
1.4 Significant contributions of the thesis
Although LFC has been widely addressed, keeping in view the changing power system envi-
ronment there is still a lot of openings to this work. The thesis has investigated and contributed
to the following areas:
• Mathematical modeling and state space forms of conventional power systems and power
systems with MSPG are presented
• New power system models with MSPG considering TCPS, AC-DC lines and restructured
power system environment are presented for LFC study
• A more practical form of optimal controllers, OFC is presented for LFC study. A simple
algorithm is presented to solve the problem using MATLAB code
• OFC is presented for the LFC of TAIPSs (i)thermal-thermal with non-reheat (ii)thermal-
thermal with reheat and (iii) hydro-thermal power plants and dynamic responses are com-
TH-1171_07610212
10
pared with FSFC
• OFC is presented for the LFC of (i) SAPS with MSPG (ii) TAIPS with MSPG and dy-
namic responses are compared with FSFC. LFC of low, medium and high head hydro
plants is also studied. Sensitivity analysis is done over a wide range of variation of pa-
rameters and load conditions. Effect of GRC and variation in regulation parameter is also
analyzed.
• OFC is presented and compared with most recent research work [1] on actual hydro plants
and improved dynamic responses are obtained
• OFC is presented for LFC of TAIPS with MSPG considering AC-DC tie lines and im-
proved dynamic responses are obtained
• OFC is presented for LFC of TAIPS with MSPG considering TCPS and improved dy-
namic responses are obtained
• OFC is presented for LFC of TAIPS with MSPG in restructured power system environ-
ment considering various possible contracts between GENCOs and DISCOs.
1.5 Organization of the thesis
The thesis is organized as follows:

Chapter 1 gives a general introduction on LFC problem. The past achievements in the
LFC literature are briefly reviewed and the main motivation and significant contributions of the
present thesis are summarized.

Chapter 2 presents mathematical modeling and LFC of power systems. The controller de-
sign equations and algorithm are presented in brief.
Chapter 3 discusses LFC of conventional power systems. The LFC of (i)Thermal-thermal
power system with non-reheat turbines, (ii) Thermal-thermal power system with reheat turbine
and (iii) Hydro-thermal power system has been studied and performance of OFC is compared
with FSFC.
TH-1171_07610212
11
Chapter 4 proposes LFC of the power systems with MSPG. The LFC of (i) Single area
power system with multi-source power generation, (ii) Two area interconnected power system
with multi-source power generation (iii) Hydro power plants has been studied.
Chapter 5 is organized in two main sections. Firstly, the LFC of power system with MSPG
considering AC-DC tie lines is presented. In the second section, LFC of power system with
MSPG considering TCPS is studied.

In chapter 6, LFC of TAIPS with MSPG in restructured power system is presented.
Finally in Chapter 7, general conclusions and suggestions for further work are documented.
TH-1171_07610212
12
C HAPTER 2
M ATHEMATICAL MODELING AND LFC OF
POWER SYSTEMS
2.1 Introduction
A detailed literature survey on LFC problem has been carried out in the previous chapter. It
is understood that the study of the response of power systems to perturbations and operational
changes is greatly assisted by mathematical models and computer simulations. Mathematical
models involving small perturbations are developed by linearization of the system around a
current operating point but the larger disturbances have to be obtained by solving non-linear
differential equations. The LFC study is basically based on the small signal analysis. The
linearized models of turbines, governors, power systems and associated equations reported in
[2, 7, 8, 10, 39] are well accepted and have been used by many researchers for LFC system
modeling in isolated and interconnected power systems [4, 13, 81, 85, 86, 95].
In the classical control methodologies, frequency response plots such as Bode and Nyquist
diagrams are usually used to obtain the desired gain and phase margins. The investigations car-
ried out using classical control approaches reveal that it often result in relatively large overshoots
and transient frequency deviation [9, 10, 34, 118]. Moreover, the settling time of the system fre-
quency deviation is comparatively long. Most recently many researchers [1,12,13,15,80,85–87]
have studied the LFC problem of hydro, thermal systems using PID controller, fuzzy controller,
decentralized controller and optimal MISO PID controller based on different algorithms and
optimization techniques. The fuzzy PI controller is known to give poor performance in system
TH-1171_07610212
13
Chapter 2 Mathematical modeling and LFC of power systems
transient response. Chang et al. [88], obtained the optimum adjustment gains of the integral
controller using GAs through performance indices ISE and ITAE. The premature convergence
of GA degrades its efficiency and reduces the search capability. Nanda et al. [15] proposed
the LFC using Bacterial Foraging-Based Optimization. PSO is developed through simulation
of bird flocking in multi-dimensional space. Like GA, PSO is also less susceptible to getting
trapped on local optimum [15, 80]. Mojtaba et al. [85] proposed a robust multi-variable model
based predictive control (MPC) for the solution of LFC in a multi-area power system. Alireza et
al. [1] studied the LFC of the hydro power system (operational in Iran) using optimal MISO PID
controller. Hasan et al. [86] presented the design of sub-optimal AGC regulators based on the
constrained feedback control strategy using the feedback of system states. Decentralized load
frequency controller is presented for the LFC of an interconnected thermal power system [13]
which uses large number of states for the controller feedback. Challa et al. [89] has presented
the analysis and design of controller for two area hydro-thermal-gas AGC system. They have
shown that for LFC study, optimal PI state feedback controller is more robust and performs
better than conventional genetic algorithm based PI controller. However, this optimal PI state
feedback controller uses all the states for feedback purpose which is practically difficult and
results in the increased complexity and cost of the controller. Rakhshani et al. [87] have ap-
plied reduced-order observer control for two-area LFC system after deregulation. In fact, one
of the main observed problems in the control of AGC systems is the limitation to access and
measurement of state variables in the real world. All these controllers discussed have relative
advantages and disadvantages. Shayeghi et al. [95] concluded in the state of art survey on LFC
that there are no rules as to when a particular technique is more suitable for the LFC problem.
Application of the optimal control theory to power system has shown that an optimal load
frequency controller can improve the dynamic stability of a power system [4]. The LFC regula-
tor design techniques using modern optimal control theory enable the power engineers to design
an optimal control system with respect to given performance criterion. Fosha and Elgerd [39]
were the first to present their work on optimal LFC regulator design using this concept. The
feasibility of an optimal LFC scheme requires the availability of all state variables for feedback.
TH-1171_07610212
14
However, these efforts seem unrealistic, since it is difficult to achieve this. Then, the problem is
to reconstruct the unavailable states from the available outputs and controls using an observer.
Due to practical limitations in the implementation of regulators based on feedback of all state
variables, suboptimal AGC regulator designs were considered [47–49].
As state feedback controllers require the transfer of information from all parts of the system
to a central control facility for processing, which for large scale inter connected power sys-
tem could prove to be prohibitive, alternative more practical forms of feedback controllers such
as output feedback and decentralized controllers have been the subject of extensive investiga-
tion. As these controllers use only a subset of the state vector for feedback purposes, they are
simpler, more practical and easy to implement than the full state feedback controllers [2, 49].
Bettayeb et. al [82] investigates the incorporation of time-weighted linear quadratic regulator
state and output feedback control design for power system dynamic stability analysis. Shahnazi
et. al [83] proposed an output feedback control with disturbance rejection for a class of non-
linear MIMO systems with unknown but bounded disturbances. An application of LQG based
load frequency controller in a competitive electricity environment is witness of the popularity

of state feedback controls [11]. Tyagi and Srivastava [84] presents the design of a decentralized
automatic generation control (AGC) scheme where the controller has been designed by appro-
priately assigning the eigen-structure of each isolated subsystem via state feedback, satisfying
the sufficient conditions for stability. Mishra et al. [119] studied LFC using linear quadratic
regulator where Kalman estimator is used to estimate the states. The estate estimation increases
the complexity and cost of the controller. Amongst various control schemes of state feedback
discussed above, the output feedback method [83, 120]seems to be more practical.
Most recent application of state feedback control (optimal control) [11,82–84,86,87,89,90,

93, 94, 119] in LFC analysis of power systems shows their popularity and superiority over other
controllers. The output feedback controller (OFC) is proposed in this thesis from pragmatic
point of view, which uses less number of states as feedback. In the output feedback method,
only the output state variables within each control area are required for feedback purpose. The
proposed OFC overcomes the drawbacks of full state feedback controller (FSFC).
TH-1171_07610212
15
This chapter presents LFC modeling, associated equations and linearized models of single
area power system (SAPS) and multi-area interconnected power systems (MAIPS) in detail.
The new LFC model for the MAIPS having multi-source power generation(MSPG) is proposed.
The fundamental and secondary LFC loops, concept of tie lines, ACEs are introduced. The
controller design steps and algorithm of the OFC are described in detail. FSFC is used for the
comparison purpose, however important equations are described in brief for the ready reference.
The important steps to solve the LFC problem using proposed OFC are described in last.
2.2 Fundamental LFC loops
The frequency of a power system depends on active power balance. A change in active power
demand at one point of a network is reflected in the entire power system by a change in fre-
quency. Therefore, system frequency provides a useful index to indicate the imbalance between
the active power generation and load. Any short-term energy imbalance will result in an instan-
taneous change in system frequency as the disturbance is initially offset by the kinetic energy of
the rotating plant. Significant loss in the generation without an adequate system response can
produce extreme frequency excursions outside the working range of the plant. Therefore LFC
plays an important role to maintain the balance between power generation and frequency [5, 6].
The real power in a power system is being controlled by controlling the mechanical power
output of the prime mover. Depending on the type of generation,the prime mover may be a
steam turbine,gas turbine, hydro-turbine or diesel engine. In the case of a steam or hydro-
turbine,mechanical power is controlled by the opening or closing of valves regulating the input
of steam or water flow into the turbine. Steam (or water) input to turbines must be continu-
ously regulated to match real power demand, failing which the machine speed will vary with
consequent change in frequency [5, 6, 10].
In addition to a primary frequency control, most of the large synchronous generators are
equipped with a supplementary frequency control loop. A schematic block diagram of a syn-
chronous generator equipped with frequency control loops is shown in Fig. 2.1 where, the speed
TH-1171_07610212
16
governor senses the change in speed (frequency) via the primary and supplementary control
loops. Very large mechanical forces are needed to position the main valve(or gate) against the
high steam (or water) pressure, and these forces are obtained via several stages of hydraulic am-
plifiers. The hydraulic amplifier provides the necessary mechanical forces to position the main
valve against the high-steam (or hydro) pressure, and the speed changer provides a steady-state
power output setting for the turbine [5, 6, 10].

The speed governor on each generating unit provides the primary speed control function, and
all generating units contribute to the overall change in generation, irrespective of the location
of the load change, using their speed governing. However, primary control action is not usually
sufficient to restore the system frequency and the supplementary control loop is required to
adjust the load reference set point through the speed-changer motor. The supplementary loop
performs a feedback via the frequency deviation and adds it to the primary control loop through
a dynamic controller. The resulting signal (∆Pc ) is used to regulate the system frequency. In
real-world power systems, the dynamic controller is usually a simple integral or proportional
integral (PI) controller [2, 5, 6, 10].

As shown in Fig. 2.1, the frequency experiences a transient change (∆ f ) following a change
in load (∆PD ). Thus, the feedback mechanism comes into play and generates an appropriate
signal for the turbine to make generation (∆Pm ), track the load and restore the system frequency
[6].
2.3 Frequency response modeling
Power systems have a highly non-linear and time-varying nature. However, for the purpose of
frequency control synthesis and analysis in the presence of load disturbances, a simple low-
order linearized model is used. In comparison with voltage and rotor angle dynamics, the
dynamics affecting frequency response are relatively slow, in the range of seconds to minutes. In
this section, a simplified frequency response model for the described schematic block diagram
in Fig. 2.1 with one generator unit is described, and then the resulting model is generalized
TH-1171_07610212
17
Pm
Rotor shaft Turbine Generator
Primary m
control loop
Pc
Speed Pv
Speed Hydraulic Valve/Gate
Controller
changer governor amplifier
Supplementary
control loop
Steam/water
f
Frequency sensor Load
PD
Figure 2.1: Schematic block diagram of a synchronous generator with basic frequency control
loops
F(s)
1/R PD(s) = Pgen(s)
- -
Pc(s) Pg(s) Pv(s) Pm(s) KPS F(s)
+
Yg (s) Yt (s)
+
(1+sTPS)
Governor Turbine generator Power system
Figure 2.2: Linear model of the primary LFC loop
for multi-area interconnected power system as described in section 2.4. A linear model of the
system with primary LFC loop is shown in Fig. 2.2 and its associated equation are presented in
this section. An increase in governor command ∆Pg results from increase in ∆Pc and a decrease
in ∆ f . Thus we can write for a small increment
1
∆Pg = ∆Pc − ∆ f (2.1)
R
where
∆Pg is small change in governor command,

∆Pc is small change in reference power setting,
and R is the regulation parameter or droop characteristics (Hertz per pu MW). R is de-
fined as the ratio of speed deviation or frequency deviation to change in valve/gate position or
TH-1171_07610212
18
power output. For example 4% R means that 4% frequency deviation causes 100% change in
valve/gate position or power output.

Laplace transformation of the equation (2.1) yields
1
∆Pg (s) = ∆Pc (s) − ∆F (s) (2.2)
R
Yg (s) , the transfer function of speed governor system is given by

∆Pv (s)
Yg (s) = (2.3)
∆Pg (s)
The speed governor system transfer function, Yg (s) depends upon the type of the governing sys-
tem used in the generating unit. A speed governing system of a steam turbine has the following
transfer function [2, 7]:
1
Yg (s) = (2.4)
1 + sTSG
where, TSG is the steam turbine speed governor time constant.
The Turbine power increment ∆Pm depends entirely upon the control valve(or gate) power
increment ∆Pv and response characteristics of the turbine.
The turbine transfer function may be given as
∆Pm (s)
Yt (s) = (2.5)
∆Pv (s)
The turbine transfer function, Yt (s) depends upon the type of turbine.
A non-reheat steam turbine has the following transfer function [2, 7]:
∆Pm (s) 1
Yt (s) = = (2.6)
∆Pv (s) 1 + sTT
where, TT is steam turbine time constant.

The generator power increment ∆Pgen = ∆PD depends entirely upon the changes ∆PD in the
load being fed from the generator, which adjusts its output so as to meet the demand changes.
These adjustments are essentially instantaneous certainly in comparison with slow changes in
turbine output, and therefore we can set
∆Pgen = ∆PD (2.7)
TH-1171_07610212
19
Power system (load and machine) transfer function is given by

KPS
Yp (s) = (2.8)
1 + sTPS
where
1
KPS = D is power system gain, Hz/pu MW,
H is the equivalent inertia constant, MW-s/MVA,
2H
TPS = fD is power system time constant, s
∂ PL 1
and D = ∂ f Prt is Equivalent system damping coefficient, pu MW/Hz.
∆F (s) can now be expressed as
∆F (s) = Yp (s) [∆Pm (s) − ∆PD (s)] (2.9)
In fact, the block diagram of a single area power system without supplementary control shown in
Fig. 2.2 is based on above equations and results. Now it is necessary that frequency deviations
must settle with zero steady state error. To accomplish this supplementary control loop must
be closed and the speed changer is manipulated in accordance with some suitable controller
action. The signal fed into the controller is referred to as area control error (ACE). The ACE in
an isolated(single area) power system may be defined as:
ACE = B∆ f (2.10)
where, B is frequency bias parameter, pu MW/Hz.
The speed changer can be commanded by a control signal ∆Pc obtained by a suitable control
action on the error signal.
∆Pc = K(s)ACE (2.11)
where, K(s) is the controller.

For an example, Let the controller be integral controller, K(s) = −KI
s then ∆Pc becomes
Z
∆Pc = −KI B∆ f dt (2.12)
The polarity of integral controller must be chosen negative so as to cause a positive frequency
error to give rise to a negative or decrease command. As long as an error remains the controller
TH-1171_07610212
20
Primary control loop
PD(s)
B Supplementary 1/R
control loop
- -
Pc(s) Pg(s) Pv(s) Pm(s) KPS F(s)
+
K(s) Yg (s) Yt (s)
+ (1+sTPS)
Controller Turbine
Governor Power system
Figure 2.3: Linear model with supplementary LFC loop
output will increase, causing the speed changer to move. The controller output, and thus the
speed changer position attains a constant value only when the frequency error has been reduced
to zero. The gain constant KI controls the rate of integration and thus speed of response of loop.
A block diagram for an isolated (single area) system with supplementary control is as shown in
Fig. 2.3.
2.4 LFC modeling of the MAIPS
In an isolated power system, regulation of tie-line power is not a control issue, and the LFC task
is limited to restore the system frequency to the specified nominal value. In order to generalize
the described model for interconnected power systems, the control area concept needs to be
used, as it is a coherent area consisting of a group of generators and loads, where all the genera-
tors respond to changes in load or speed-changer settings, in unison. The frequency is assumed
to be the same at all points of a control area. A multi-area power system comprises areas that are
interconnected by high voltage transmission lines or tie lines. The trend of frequency measured
in each control area is an indicator of the trend of the mismatch power in the interconnection
and not in the control area alone. The LFC system in each control area of the multi-area inter-
connected power systems (MAIPS) should control the interchange power with the other control
areas as well as its local frequency [2, 5, 6, 10]. Therefore, the described dynamic LFC system
model shown in Fig. 2.3 must be modified by taking into account the tie-line power signal. For
TH-1171_07610212
21
Ptie, 12 Ptie, 1N
Control area 2 Control area 1 Control area N
( f2 , , V2 ) ( f1 , 1 , V1 ) ( fN , , VN )
2 X 12 X 1N N
Ptie, 13
Ptie, 14
X 13
X 14
Control area 3 Control area 4
( f3 , 3 , V3 ) ( f4 , 4 , V4 )
Figure 2.4: N-control areas power system
this purpose, consider Fig. 2.4, which shows a power system with N-control areas. In normal
operation the tie line power flow from control area 1 to control area 2 is given by
|V1 | |V2 |
Ptie,12 = sin(δ1 − δ2 ) (2.13)
X12
where X12 is the tie-line reactance between control areas 1 and 2,

and δ1 and δ2 are the angles of end voltages V1 and V2 respectively.
For small deviations in the angles δ1 and δ2 the tie line power changes with
|V1 | |V2 |
∆Ptie,12 = cos(δ1 − δ2 )(∆δ1 − ∆δ2 ) (2.14)
X12
Synchronizing coefficient T12 of a line is defined by

|V1 | |V2 |
T12 = cos(δ1 − δ2 ) (2.15)
X12
The tie line power deviation then takes on the form
∆Ptie,12 = T12 (∆δ1 − ∆δ2 ) (2.16)
The frequency deviation ∆ f is related to the reference angle ∆δ by the formula

Z t
∆δ = 2 π ∆ f dt (2.17)
0
By expressing the tie line power deviations in term of ∆ f , we get

Z t Z t
∆Ptie,12 = 2π T12 ∆ f1 dt − ∆ f2 dt (2.18)
0 0
TH-1171_07610212
22
Laplace transformation of the equation (2.18) yields

2π T12
∆Ptie,12 (s) = (∆F1 (s) − ∆F2 (s)) (2.19)
s
Similarly the incremental tie line power flow from control area 2 to area 1 is given by
Z t Z t
∆Ptie,21 = 2π T21 ∆ f2 dt − ∆ f1 dt (2.20)
0 0
Taking the Laplace transform of equation (2.20)

2π T21
∆Ptie,21 (s) = (∆F2 (s) − ∆F1 (s)) (2.21)
s
where,
|V2 | |V1 |
T21 = cos(δ2 − δ1 ) (2.22)
X12
If incremental powers are expressed in pu, then
Prt1
T21 = − T12 = a12 T12 (2.23)
Prt2
where
Prt1 and Prt2 are the rated power of control area 1 and 2 respectively,
and
a12 is the control area capacity ratio, defined as: a12 = − PPrt1
rt2
If losses are neglected and equal area ratings (Prt1 = Prt2) are taken
∆Ptie,21 = a12 ∆Ptie,12 = −∆Ptie,12 (2.24)
similarly, the tie line power change between areas 1 and 3 is given by
2π T13
∆Ptie,13 (s) = (∆F1 (s) − ∆F3 (s)) (2.25)
s
and the tie line power change between control areas 1 and 4 is given by
2π T14
∆Ptie,14 (s) = (∆F1 (s) − ∆F4 (s)) (2.26)
s
Considering equations (2.19), (2.25) and (2.26), the total tie line power change between control
area 1 and the other control three areas 2, 3 and 4 can be calculated as
!
2π
∆Ptie,1 (s) = ∆Ptie,12 (s) + ∆Ptie,13 (s) + ∆Ptie,14(s) =
s ∑ T1 j ∆F1 (s) − ∑ T1 j ∆Fj (s) (2.27)
j=2,3,4 j=2,3,4
TH-1171_07610212
23
Fi (s)
"T
j 1
ij
j !i
+
P tie,i(s) !"#
-
N
# T "F (s)
j 1
ij j
j !i
Figure 2.5: Block diagram representation for tie line power change of control area-i in an inter-
connected power system with N-control areas
Similarly, for N-control areas shown in Fig. 2.4, the total tie line power change between control
area-i and all other control areas is
 
N
2π  N N
∆Ptie,i (s) = ∑ ∆Ptie,i j (s) =  ∑ Ti j ∆Fi (s) − ∑ Ti j ∆Fj (s) (2.28)

j=1 s j=1 j=1
j6=i j6=i j6=i
Equation (2.28) is represented in the form of a block diagram in Fig. 2.5. The effect of changing
the tie line power for an area is equivalent to changing the load of that area. Therefore, the ∆Ptie,i
must be added to the mechanical power change ∆Pmi and control area load change ∆PDi using
an appropriate sign [2, 5, 6, 10]. A combination of block diagrams Figs. 2.3 and 2.5 creates a
simplified block diagram for control area-i in an N-control area interconnected power system
shown in Fig. 2.6. The next point to consider is the supplementary control loop in the presence
of a tie line. In the case of an isolated control area, this loop is performed by a feedback
from a control area frequency deviation through a simple dynamic controller as shown in Fig.
2.3. In a MAIPS, in addition to regulating area frequency, the supplementary control should
maintain the tie line power interchange with neighboring areas at scheduled values. This is
generally accomplished by adding a tie line power flow deviation to the frequency deviation in
the supplementary feedback loop. A linear combination of frequency and tie line power changes
TH-1171_07610212
24
PDi (s)
Bi 1/Ri
+ - -
Pgi(s) Pvi (s) Pmi (s) KPSi Fi (s)
ACEi Pci (s) +
Ki (s) Ygi (s) Yti (s)
+ (1+sTPSi)
+ Controller Turbine -
Governor Power system
N
"T
j 1
ij
j !i
P tie,i +
!"#
-
N
# T "F (s)
j 1
ij j
j !i
Figure 2.6: Control area-i of an interconnected power system with N number of control areas
for control area-i, is known as the ACE and can be given as:
ACEi = ∆Ptie,i + Bi ∆ fi (2.29)
Bi is the frequency bias parameter of ith control area. Bi is assumed equal to the frequency bias
factor(βi ) [2, 5, 6, 10].
1
B i = β i = Di + (2.30)
Ri
The block diagram shown in Fig. 2.6 illustrates how supplementary control is implemented
using equation (2.29). The effects of local load changes and interface with other control areas
are properly considered as two input signals. Each control area monitors its own tie line power
flow and frequency at the area control centre. The ACE signal is computed and allocated to
the controller K(s). Finally, the speed changer of ith control area can be commanded by a
corresponding control signal ∆Pci . Therefore, it is expected that the supplementary control
shown in Fig. 2.6 can ideally meet the basic LFC objectives and maintain area frequency and
tie line interchange at scheduled values.
TH-1171_07610212
25
Primary control
1/R1i 1/R2i 1/Rki

Governor Turbine
Bi - Pv1i(s) Pm1i(s)
Pcc1i
1i(s)
(s + Yt1i (s)
PDi (s)
Yg1i(s) 1i
Supplementary
Governor Turbine +
+ Control -
- KPSi Fi (s)
+ Pv2i(s) Pm2i(s) +
1/s Pcc2i
2i(s)
(s Yt2i (s)
Yg2i(s) 2i
ACEi ACEi (1+sTPSi)

+ + -
Power system
N
Governor Turbine
- Pvki(s) Pmki(s)
"T ij
Pcckiki(s)
(s + j 1
Ygki(s) Ytki(s) ki j !i
+
!"#
P tie,i
-
N
# T "F (s)
j 1
ij j
j !i
Figure 2.7: Control area-i of MAIPS with MSPG and N-number of control areas
2.5 LFC modeling of MAIPS with MSPG
Since the 1970s, the described LFC scheme and power system model in section 2.4 is widely
used by researchers for the LFC analysis and synthesis. The far reaching restructuring of the
power system industry, application of new technology and concomitant new concepts of oper-
ation requires an evaluation and re-examination of this scheme, which is already designed to
operate with large and central generating facilities to find ways to maintain, and possibly im-
prove, their efficiency and reliability. Most of the researchers considered either thermal or hydro
generating units in a control area. In a real situation, control area may have variety of sources
of generations such as hydro, thermal, gas, nuclear, solar, wind etc and the control area having
different sources of power generation represented by an equivalent of thermal or hydro unit

dynamics only may not result in realistic design of LFC control [6]. Furthermore, in the new
power system environment, generators may or may not contribute in the LFC task and contribu-
tion factors are not the same for all participant generators [1, 5, 6, 17, 89, 116]. Keeping in view
the present power scenario, combination of multi-source generators in a control area with their
TH-1171_07610212
26
corresponding generation contribution factors is more realistic for the study of LFC [89, 116].
In a competitive environment, generation contribution factors are actually time-dependent vari-

ables and must be computed dynamically based on bid prices, availability, congestion problems,
costs and other related issues. In order to consider the variety of generation dynamics and their
contribution rate in the LFC action, the dynamic model of control area-i in Fig. 2.6, can be
modified to that shown in Fig. 2.7. From the Fig. 2.7, the total incremental generation change
in turbine power outputs of ith area may be given as:
n̄
∆Pmi (s) = ∑ ∆Pmki(s) (2.31)
k=1
where, n̄ is number of generating units(GENCOs) in the ith control area.

Let the generation contribution factor of kth generating unit in the ith area be defined as αki .
The ∆Pmki may now be expressed in terms of ∆Pmi (s) and αki as:
∆Pmki (s) = αki ∆Pmi (s) (2.32)
From the equation (2.32), it is obvious that ,the sum of generation contribution factors in a
control area [5, 6, 89, 116] is equal to 1.

n̄
∑ αki = 1 (2.33)
k=1
where, 0 ≤ αki ≤ 1.
Considering the effect of primary and supplementary controls, the system frequency can be
obtained as:
!
n̄
KPSi
∆Fi (s) =
(1 + sTPSi ) ∑ ∆Pmki(s) − ∆PDi(s) − ∆Ptie,i(s) (2.34)
k=1
and ∆Pmki (s) can be expressed [6, 17, 116] as:
∆Fi (s)

∆Pmki (s) = Ygki (s)Ytki (s)αki ∆Pcki (s) − (2.35)
Rki
where, Ygki (s) and Ytki (s) are the transfer functions of governing system and turbines of kth
generating unit in the ith control area, respectively.
TH-1171_07610212
27
using equations(2.35), the equation(2.34) results as:

" #
n̄
∆Fi (s)

KPSi
∆Fi (s) = ∑ Ygki (s)Ytki(s)αki ∆Pcki(s) − Rki − ∆PDi(s) − ∆Ptie,i(s) (2.36)
(1 + sTPSi ) k=1
It is assumed that for ith control area, Rki = Ri , for (k = 1 to n̄) and SLP is given in the control
area. After knowing the generating units, corresponding Ygki (s) and Ytki (s) can be substituted
into equation (2.36) to obtain the frequency response of the ith control area. The LFC study of
this kind of model is presented in the chapter 4.
2.6 Controller Design
Probably the most important contribution, optimal control theory has made to the control engi-
neer is the ability to handle a large multivariate control problem with ease. The engineer has
only to represent the control system in state variable form and specify the desired performance
mathematically in terms of a cost to be minimized. A unique or best controller in the sense of
minimizing the cost may be generated by applying well proven theories and techniques. Most
recent application of state feedback controllers(optimal control) [11, 82–84, 86, 87, 89, 90, 93,
94, 119] in the LFC analysis of power systems shows their popularity and superiority over other
controllers.
In modern control theory approach, control signals are generated by a linear combination
of all the system states (full state feedback approach) or a linear combination of states to be
controlled/measurable states (output feedback approach) [2, 7, 120]. The OFC is proposed to
improve the LFC problem from practical point of view. Practically, it is very difficult and of-
ten expensive to measure and to have readily available information about all the states in most
of the large power systems. Moreover, the implementation of optimal AGC regulator requires
monitoring of all the state variables of the system or state reconstruction, which may be undesir-
able from cost and complexity considerations [86]. Usually reduced number of state variables
or a linear combination thereof is available. The proposed OFC uses less number of feedback
states as compared to FSFC, thereby reduces number of sensors, cost and complexities. In this
section, FSFC is described in brief for the ready reference and the proposed OFC is explained
TH-1171_07610212
28
in a detail with algorithm.
The statement of the problem in an optimal control theory, may be described in the following
state-space form: [1, 2, 7, 39, 119, 120]:
ẋ = Ãx + B̃u (2.37)
and
y = C̃x (2.38)
where
x is a state vector of the dimension n × 1, n is no. of state variables
u is a control vector of the dimension m × 1, m is no. of control variables

y is an output vector of the dimension p × 1, p is no. of output variables
Ã, B̃ and C̃ are constant matrices with dimensions of n × n, n × m and p × n, respectively.
The performance of the system is specified in terms of a performance index or cost function(J):
Z∞
1
xT Q̃x + uT R̃u dt

J= (2.39)
2
0
which is minimized for obtaining parameters of an optimal controller. In the equation (2.39), Q̃
is n ×n, symmetric positive semi-definite state cost weighting matrix and R̃ is m ×m, symmetric
positive semi-definite control cost weighting matrix. The elements of the matrices Q̃ and R̃ may
be chosen as per the designer’s choice.
The optimal controller law for full state feedback can be defined by [2]
u = −K̄x (2.40)
The constant gain matrix K̄ of the dimension m × n, is obtained from the solution of the matrix
Riccati equation
ÃT P + PÃ − PB̃R̃−1 B̃T P + Q̃ = 0 (2.41)
K̄ = R̃−1 B̃T P (2.42)
For stability, all the eigenvalues of the matrix (Ã − B̃K̄) should have negative real parts. From
TH-1171_07610212
29
equations (2.41) and (2.42), the optimal setting of FSFC gains(K̄) is obtained. The solution of
Riccati equation (2.41) is obtained using the MATLAB function care.m [121, 122].
Let the output feedback control law be defined as
u = −K̃y (2.43)
where K̃ is OFC gain matrix of dimension (m × p).

In the optimal control scheme the control inputs are generated by means of feedbacks from
the output states with feedback constants to be determined in accordance with optimality cri-
terion. Using equations(2.38) and (2.43), the linear model given by equation (2.37) can be
arranged as
ẋ = (Ã − B̃K̃C̃)x = ÃC x (2.44)
The performance index given in equation (2.39) can be rewritten as

Z∞
1
xT Q̃ + C̃T K̃ T R̃K̃C̃ x dt

J= (2.45)
2
0
The control problem is now to design the gain matrix K̃ so that J is minimized subject to the
following dynamical constraint:
ẋ = (Ã − B̃K̃C̃)x (2.46)
This dynamical optimization problem may be converted into an equivalent static one that is eas-
ier to solve. After optimization and simplification, the following optimal gain design equations
are obtained [120]:
0 = ÃTc P̃ + P̃Ãc + C̃T K̃ T R̃K̃C̃ + Q̃ (2.47)
0 = Ãc S̃ + S̃ÃTc + X (2.48)
K̃ = R̃−1 B̃T P̃S̃C̃T (C̃S̃C̃T )−1 (2.49)
where
X = E x(0)xT (0)

If initial states are assumed to be uniformly distributed on the unit sphere, then X = I, where
TH-1171_07610212
30
X is n × n, symmetric matrix and I is an identity matrix. In many applications x(0) may not be
known, this dependence is typical of output feedback design. It is usual to sidestep this problem
by minimizing not the performance index [120] but its expected value (E {J}) ,
1 1
E {J} = E xT (0)P̃x(0) = tr(P̃X ) (2.50)
2 2
The optimal cost can be given by

1
J0 = tr(P̃X ) (2.51)
2
The equations(2.47) and (2.48) are Lyapunov equations and the equation (2.49) is an equation
for the gain K̃. To obtain the output feedback gain K̃ minimizing the J0 , these three coupled
equations may be solved simultaneously by iterative techniques [120]. Algorithm to solve these
three coupled equations is presented in the section 2.6.1.
2.6.1 Algorithm for OFC
The algorithm for the proposed optimal OFC is as follows:
1.Initialize
Set j=0
set initial value of the gain K̃0 so that Ãc,0 = (Ã − B̃K̃0C̃) is asymptotically stable
2. jth iteration
Set Ãc, j = (Ã − B̃K̃ jC̃)

solve Lyapunov equations for P̃j and S̃ j
0 = ÃTc, j P̃j + P̃j Ãc, j + C̃T K̃ Tj R̃K̃ jC̃ + Q̃
0 = Ãc, j S̃ j + S̃ j ÃTc, j + X
solve for K̃ j , where j 6= 0
K̃ j = R̃−1 B̃T P̃j S̃ jC̃T (C̃S̃ jC̃T )−1

Set
J0, j = 12 tr(P̃j X )
Evaluate
∆K̃ = R̃−1 B̃T P̃S̃C̃T (C̃S̃C̃T )−1 − K̃ j
TH-1171_07610212
31
K̃ j+1 = K̃ j + σ ∆K̃
where σ is chosen so that Ãc, j+1 is asymptotically stable

J0, j+1 = 12 tr(P̃j+1X ) ≤ J0, j
If the values of J j+1 and J j are very close to each other go to step 3
otherwise set j = j + 1 and go to step 2
3.Terminate
Set K̃ = K̃ j+1
and J0 = J0, j+1
Stop
MATLAB code is developed based on the above algorithm to get the optimal value of OFC
gains(K̃).The Lyapunov equations are solved using MATLAB function lyap.m [121, 122].
2.6.2 Selection of Q̃ and R̃ matrices
The elements of matrices Q̃ and R̃ may be chosen as per the designer’s choice [120]. To obtain
the desired dynamic response and for the dynamic correction of ACE of LFC systems discussed
in this thesis, the following design criterion are considered to obtain the matrices Q̃ and R̃
[2, 90, 120, 123]:
1. Excursions of ACEs about their steady values are minimized.
2. Excursions of ∫ ACEdt about the steady values are minimized.
3. Excursions of control vector about their steady values are minimized.
2.7 Important steps to solve LFC problem
1. Obtain the state equations for the LFC model of the power systems and represent in the
state space form.
2. Input the system simulation parameters and data.
3. Compute matrices Ã, B̃ and C̃.
TH-1171_07610212
32
4. Compute matrices Q̃ and R̃ as detailed in 2.6.2 .
5. Obtain the optimal controller gains as described in the section 2.6.
6. Simulate the power system with optimal controllers using MATLAB simulation.
7. Obtain the dynamic responses of the power system following SLPs.
2.8 State-space model of the power systems
State-space model of a LFC dynamical system is a useful representation for the application of
the modern/robust control theory. Using appropriate definitions and state variables, the state-
space realization of the isolated and interconnected power systems is obtained in the form of
equation (2.52) and (2.38) [1, 2, 7, 10, 13, 39, 48, 90, 119].
ẋ = Ãx + B̃u + F̃w (2.52)
where,
F̃ is a disturbance distribution matrix of the dimension n × d,
w is a disturbance vector of the dimension d × 1,

d, is the number of disturbance inputs, and all others terms are same as described in the
equations (2.37) and (2.38).
The term (F̃w) represents the disturbance, in way of load demand, to the power system.
On the comparison, this power system model described in equation (2.52) is not in the form
of the equation (2.37)as desired, for two reasons. In the optimal control theory used, there is
no (F̃w) term, and second, the cost function requires the states to be driven to zero for it to
have a minimum value. For a SLP in control area-i, the steady state frequency deviations in
each control area are required to be zero. But the increased generation in control area-i will by
necessity in steady state equal the increased load demand of the area, a non zero quantity, refer
Fig. 2.6:
∆Pmi = ∆PDi (2.53)
TH-1171_07610212
33
Some of the other states will also be non-zero [2,8,39,49]. Fosha et al. [39] are probably the first
who addressed this dilemma. A simpler way of addressing this dilemma [39, 49] is to redefine
the state variables and control variables in terms of their corresponding steady state values.
The ith state variable xi may be redefined as:
∆
x1i = xi − xiss (2.54)
where i = 1, 2, ...., n, and xiss is the steady state value of ith state variable.
Similarly ith control variables may be redefined as:
∆
u1i = ui − uiss (2.55)
where i = 1, 2, ...., m, and uiss is the steady state value of ith control variable.
This change of variables puts the system in the form
ẋ1 = Ãx1 + B̃u1

(2.56)
x1 (0) = −xss
By redefining the states and control variables in terms of their steady state values, only the
reference position of the system is shifted [2, 8, 39, 49]. The superscript 1 has been dropped
[2,8,39,49] to prevent unnecessary notation problems. The matrices Ã and B̃ remain unchanged
and the equation (2.56) has the form of the equation (2.37).
2.9 Summary
Mathematical modeling of the SAPS, MAIPS and MAIPS with MSPG are described for LFC
study. Keeping in view the new power system environment, the concept of MSPG in control
area is introduced. In addition to LFC models of SAPS and MAIPS, a new model of MAIPS
with MSPG is presented for LFC study. The general form of state-space model of the power
systems is explained. The OFC proposed with pragmatic point of view is introduced . The
control system design equations and algorithms have been presented.
TH-1171_07610212
34
C HAPTER 3
LFC OF CONVENTIONAL POWER SYSTEMS
3.1 Introduction
LFC modeling and control design of power systems have been presented in detail in chapter 2.
Recently, many researchers studied the LFC of conventional power systems considering either
thermal-thermal or hydro-thermal systems [9, 12, 13, 107, 124]. Most of the research papers re-
ported in the area of LFC considered the two area interconnected power system with non-reheat
type thermal systems. An attempt is made in this chapter to study the LFC of interconnected
power systems including reheat type turbines and hydro turbines. The LFC of thermal-thermal
with non reheat turbine, thermal-thermal considering reheat turbine and hydro-thermal system
is studied using OFC and performance of the proposed OFC has been compared with FSFC.
3.2 TAIPS considering non-reheat turbines
A TAIPS shown in the Fig. 3.1 is considered for the LFC analysis in this section. Each control
area of this power system comprises thermal unit with non-reheat turbine. In accordance with
modern control terminology ∆Pc1 and ∆Pc2 will be referred to as control inputs u1 and u2 . In the
conventional approach u1 and u2 were provided by the integral of ACEs [2, 5, 6, 10]. In modern
control theory approach u1 and u2 will be created by a linear combination of all the system
states (full state feedback approach) or output/measurable states(output feedback approach).
For formulating the state variable model, the conventional secondary feedback loops are opened
and each time constant is represented by a separate block as shown in Fig. 3.1. State variables
TH-1171_07610212
35
Chapter 3 LFC of conventional power systems
Control area-1
B1 1/R1 PD1(s)
+ - -
ACE1 Pg1(s) 1 Pv1 (s) 1 Pm1(s) KPS1 F1(s)
ACE1 Pc1 (s) +
1/s (1+sTSG1 ) (1+sTT1)
x8 x3 x2 + x1
u1 (1+sTPS1)
+ Governor -
Turbine Power system
T12
+
P tie,1
!"#
x7
a12 -
a12
T12
+ Pm2(s) - KPS2
ACE2 ACE2 Pc2 (s) + Pg2(s) 1 Pv2 (s) 1 F2(s)
1/s
(1+sTSG2 ) x6 (1+sTT2) x5 +
x9 u2 (1+sTPS2) x4
+ - -
Turbine Power system
Governor
B2 1/R2 PD2(s)
Control area-2
Figure 3.1: TAIPS with non-reheat turbines
are defined as the outputs of all the blocks having either an integrator or a time constant.
3.2.1 State-space Model
The system shown in Fig. 3.1 has nine state variables where x1 = ∆ f1 , x4 = ∆ f2 , x2 = ∆Pm1 ,
x5 = ∆Pm2 , x8 = ∫ ACE1 dt, x9 = ∫ ACE2 dt. The disturbance vector is defined as w1 = ∆PD1
and w2 = ∆PD2 . After defining all the states as shown in Fig. 3.1, the following differential
d
equations are obtained replacing, s by dt .
1 KPS1 KPS1 KPS1

ẋ1 = − x1 + x2 − x7 − w1 (3.1)
TPS1 TPS1 TPS1 TPS1
1 1
ẋ2 = − x2 + x3 (3.2)
TT 1 TT 1
1 1 1
ẋ3 = − x1 − x3 + u1 (3.3)
R1 TSG1 TSG1 TSG1
1 KPS2 a12 KPS2 KPS2

ẋ4 = − x4 + x5 − x7 − w2 (3.4)
TPS2 TPS2 TPS2 TPS2
TH-1171_07610212
36
1 1
ẋ5 = − x5 + x6 (3.5)
TT 2 TT 2
1 1 1
ẋ6 = − x4 − x6 + u2 (3.6)
R2 TSG2 TSG2 TSG2
ẋ7 = 2π T12 x1 − 2π T12 x4 (3.7)
ẋ8 = B1 x1 + x7 (3.8)
ẋ9 = B2 x4 + a12 x7 (3.9)
The state equations (3.1)-(3.9) can be organized in the state-space form as described by the
equations (2.52) and (2.38) in the chapter 2. For the power system model shown in Fig. 3.1 the
associated vectors and matrices are as follows:
state vector, x = [x1 x2 .......x9]T ,

control vector, u = [u1 u2 ]T = [∆Pc1 ∆Pc2 ]T
disturbance vector, w = [w1 w2 ]T = [∆PD1 ∆PD2 ]T
 
KPS1
 − TPS1
1
TPS1 0 0 0 0 − KTPS1
PS1
0 0
 

 0 − T1T 1 T1T 1 0 0 0 0 0 0
 
 − R1 T1SG1 0 − TSG1 1
0 0 0 0 0 0
 
 
KPS2
1
−a12 KTPS2
 

 0 0 0 − TPS2 TPS2 0 PS2
0 0
 
Ã = 
 0 0 0 0 − T1T 2 T1T 2 0 0 0
 
0 0 0 − R2 T1SG2 0 − TSG2 1
0 0 0
 

 
 2π T12 −2π T12
 
 0 0 0 0 0 0 0
 
 B 0 0 0 0 0 1 0 0
 1 
 
0 0 0 B2 0 0 a12 0 0
 
1
 0 0 TSG1 0 0 0 0 0 0 
B̃T =  
1
0 0 0 0 0 TSG2 0 0 0
TH-1171_07610212
37
and
 
− KTPS1
PS1
0 0 0 0 0 0 0 0
F̃ T = 


0 0 0 − KTPS2
PS2
0 0 0 0 0
The matrix C̃ is defined as

 
1 0 0 0 0 0 0 0 0
 
0 0 0 1 0 0 0 0 0
C̃ = 
 

 0 0 0 0 0 0 0 1 0 
0 0 0 0 0 0 0 0 1
The performance index J is given in equation(2.39), the matrices Q̃ and R̃ are defined for this
problem using the design considerations given in the section 2.6.2. Excursions of ACEs (x7 +
B1 x1 , a12 x7 + B2 x4 ), ∫ ACEdt(x8 , x9 ) and control vectors(u1, u2 ) about their steady values are
minimized and J can be written as:
Z∞
1 h i
(x7 + B1 x1 )2 + (a12 x7 + B2 x4 )2 + x28 + x29 + u21 + u22 dt

J= (3.10)
2
0
From the equation (3.10), the matrices Q̃ and R̃ can be recognized as:
 2 
B1 0 0 0 0 0 B1 0 0
 0 0 0 0 0 0 0 0 0 
 
0 0 0 0 0 0 0 0 0
 
 
B22
 0 0 0 0 0 a12 B2 0 0

 
Q̃ = 
 
0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 
 
1+a212

 B1 0 0 a12 B2 0 0 0 0 
 
 0 0 0 0 0 0 0 1 0 
0 0 0 0 0 0 0 0 1

1 0
R̃ =
0 1
3.2.2 Simulation results and discussion
The procedure described in the section 2.7 is followed. The system parameters and data are
given in Table 3.1. The optimum gains of OFC and FSC are obtained. The computer simulations
are carried out with the optimum controller gain settings. MATLAB control system toolbox
[121] is used to simulate the power system and to obtain dynamic responses of the system for
TH-1171_07610212
38
1% SLP in the control area 1.
Table. 3.1: Simulation parameters of TAIPS considering non-reheat turbines
Parameter Value
Prt1 = Prt2 2000 MW
f 60Hz
H1 = H2 5 MW-s/MVA
D1 = D2 0.00833 puMW/Hz
TSG1 = TSG2 0.08s
TT 1 = TT 2 0.5s
R1 = R2 2.4 Hz/puMW
TPS1 = TPS2 20s
KPS1 = KPS2 120 s
B1 = B2 0.425
a12 -1
2π T12 0.215
The following computational results are obtained:

 
−0.0500 6.0000 0 0 0 0 −6.0000 0 0
 0 −2.0000 2.0000 0 0 0 0 0 0 
 
−5.2083 0 −12.5000 0 0 0 0 0 0
 
 
 0 0 0 −0.0500 6.0000 0 6.0000 0 0

 
Ã = 
 
0 0 0 0 −2.0000 2.0000 0 0 0 
 
 0 0 0 −5.2083 0 −12.5000 0 0 0 
 
 0.2150 0 0 −0.2150 0 0 0 0 0 
 
 0.4250 0 0 0 0 0 1.0000 0 0 
0 0 0 0.4250 0 0 −1.0000 0 0

T 0 0 12.5000 0 0 0 0 0 0
B̃ =
0 0 0 0 0 12.5000 0 0 0
 
0.1806 0 0 0 0 0 0.4250 0 0
 0 0 0 0 0 0 0 0 0 
 
0 0 0 0 0 0 0 0 0
 
 
 0 0 0 0.1806 0 0 −0.4250 0 0

 
Q̃ = 
 
0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 
 
 0.4250 0 0 −0.4250 0 0 2.0000 0 0 
 
 0 0 0 0 0 0 0 1.0000 0 
0 0 0 0 0 0 0 0 1.0000

1 0
R̃ =
0 1
TH-1171_07610212
39
Table. 3.2: Dynamic response comparison in terms of peak overshoot (OS)
Peak OS of ∆ f1 (Hz) Peak OS of ∆ f2 (Hz) Peak OS of ∆Ptie,1 (pu MW)

FSFC -0.02785 -0.01917 -0.004937
OFC -0.02582 -0.008737 -0.004061
% reduction in Peak OS 7.28 54.42 17.74
Table. 3.3: Dynamic response comparison in terms of ST
ST of ∆ f1 (s) ST of ∆ f2 (s) ST of ∆Ptie,1 (s)

FSFC 24.55 26 23.84
OFC 17.73 18 16.23
% reduction in ST 27.78 30.76 31.92
The optimum gains of FSFC K̄ and OFC K̃ are obtained as:

0.4048 0.9424 0.1408 −0.0320 −0.0683 −0.0096 −0.2615 1.0000 0.0000
K̄ =
−0.0320 −0.0683 −0.0096 0.4048 0.9424 0.1408 0.2615 0.0000 1.0000

0.1120 0.0319 0.5646 0.0928
K̃ =
0.0319 0.1120 0.0928 0.5646
Dynamic responses of the system are obtained for 1% SLP in the control area 1. The dynamic
responses are shown in Fig. 3.2. It is observed that the OFC gives better dynamic responses
having relatively smaller peak overshoot and lesser settling time with zero steady state error
as compared to that of FSFC. The quantitative comparison is made in Table 3.2 and 3.3 where
percentage improvement in the peak OS of ∆ f1 , ∆ f2 and ∆Ptie,1 is 7.28, 54.42, and 17.74,
respectively and percentage improvement in the settling time of ∆ f1 , ∆ f2 and ∆Ptie,1 is 27.78,
30.76, and 31.92, respectively.
3.3 TAIPS considering reheat and non-reheat turbines
In this section LFC of an interconnected power system as shown in the Fig. 3.3 is presented.
The reheat turbine is considered in control area 1. The procedure described in the section2.7
is adopted. The system parameters and data are given in Table 3.4. The power system has
10 state variables. State variables x1 , x5 , x9 and x10 are taken as output feedback states. The
optimum gains of OFC and FSFC are obtained. The computer simulations are carried out with
TH-1171_07610212
40
0.015
FSFC
0.01 OFC
Frequency deviation,area−1(Hz)
0.005
−0.005
−0.01
−0.015
−0.02
−0.025
−0.03
0 5 10 15 20 25 30
time, s
(a) Frequency deviation response of control area 1
0.015
FSFC
OFC
0.01
0.005
−0.005
−0.01
−0.015
−0.02
0 5 10 15 20 25 30
time, s
(b) Frequency deviation response of control area 2

−3
x 10
2
FSFC
OFC
Tie line power deviation (pu MW)
−1
−2
−3
−4
−5
0 5 10 15 20 25 30
time, s
(c) Tie line power deviation response
Figure 3.2: Dynamic responses of TAIPS with non-reheat turbines
TH-1171_07610212
41
Control area-1
B1 1/R1 PD1(s)
+ - -
ACE1 + Pg1(s) 1 Pv1 (s) Pm1(s) KPS1 F1(s)
ACE1 Pc1 (s) 1 (1+sKR1TR1)
1/s (1+sTSG1 )
x9 u1 x4 (1+sTT1)
x3 (1+sTR1) x2 + x1
(1+sTPS1)
+ -
Governor
Reheat Turbine Power system
T12
+
P tie,1
!"#
x8
a12 -
a12
T12
+ - KPS2
ACE2 ACE2 Pc2 (s) + Pg2(s) 1 Pv2 (s) 1 Pm2(s) F2(s)
1/s
(1+sTSG2 ) x7 (1+sTT2) x6 + x5
x10 u2 (1+sTPS2)
+ - -
Non reheat turbine Power system
Governor
B2 1/R2 PD2(s)
Control area-2
Figure 3.3: TAIPS considering reheat and non-reheat turbines
the optimum controller gain settings. MATLAB control system toolbox [121] is used to simulate
the power system and to obtain dynamic responses of the system for 1% SLP in the control area
1.

 
−0.0500 6.0000 0 0 0 0 0 −6.0000 0 0
 0 −3.3333 3.3333 0 0 0 0 0 0 0

 
 
 −2.6042 0 −0.1000 −6.1500 0 0 0 0 0 0 
 
 −5.2083 0 0 −12.5000 0 0 0 0 0 0 
 
 0 0 0 0 −0.0500 6.0000 0 6.0000 0 0 
Ã =  

 0 0 0 0 0 −3.3333 3.3333 0 0 0 

−12.5000 −12.5000
 0 0 0 0 0 0 0 0

 
 
 0.5450 0 0 0 −0.5450 0 0 0 0 0 
 
 0.4250 0 0 0 0 0 0 1.0000 0 0 
0 0 0 0 0.4250 0 0 −1.0000 0 0

T 0 0 0.5000 1.0000 0 0 0 0 0 0
B̃ =
0 0 0 0 0 0 1.0000 0 0 0
TH-1171_07610212
42
Table. 3.4: Simulation parameters of TAIPS considering reheat and non-reheat turbines
Parameter Value
Prt1 = Prt2 2000 MW
f 60Hz
H1 = H2 5 MW-s/MVA
D1 = D2 0.00833 puMW/Hz
TSG1 = TSG2 0.08s
TT 1 = TT 2 0.3s
KR1 0.5
TR1 10s
R1 = R2 2.4 Hz/puMW
KPS1 = KPS2 120 s
TPS1 = TPS2 20s
B1 = B2 0.425
a12 -1
2π T12 0.545
and the matrix C̃ is defined as

 
1 0 0 0 0 0 0 0 0 0
 
0 0 0 0 1 0 0 0 0 0
C̃ = 
 

 0 0 0 0 0 0 0 0 1 0 
0 0 0 0 0 0 0 0 0 1
 
0.1806 0 0 0 0 0 0 0.4250 0 0
 0 0 0 0 0 0 0 0 0 0

 
 
 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0.1806 0 0 −0.4250 0 0 
Q̃ =  

 0 0 0 0 0 0 0 0 0 0 

 0 0 0 0 0 0 0 0 0 0

 
 
 0.4250 0 0 0 −0.4250 0 0 2.0000 0 0 
 
 0 0 0 00 0 0 0 1.0000 0 
0 0 0 0 0 0 0 0 0 1.0000

1 0
R̃ =
0 1
The optimum gains of FSFC K̄ and OFC K̃ are obtained as:

0.2106 0.3836 7.0445 −3.4432 −0.0093 −0.0259 −0.0075 −0.0751 0.9649 −0.2627
K̄ =
−0.0313 −0.0414 −0.6459 0.3155 0.1080 0.2079 0.0553 −0.2854 0.2627 0.9649

0.0533 −0.0948 0.8087 −0.2321
K̃ =
0.0717 −0.0424 0.2888 0.9435
TH-1171_07610212
43
Dynamic responses of the system are obtained for 1% SLP in the control area 1. The dynamic
responses are shown in Fig. 3.4. It is observed that the OFC gives better dynamic responses
having relatively smaller peak overshoot and lesser settling time with zero steady state error
as compared to that of FSFC. The quantitative comparison is made in Table 3.5 and 3.6 where
percentage improvement in the peak OS of ∆ f2 and ∆Ptie,1 is 6.84 and 24.08, respectively and
percentage improvement in the settling time of ∆ f1 , ∆ f2 and ∆Ptie,1 is 7.96, 3.47 and 16.00,
respectively.

FSFC -0.02487 -0.01695 -0.01034
OFC -0.02487 -0.01579 -0.00785
% reduction in Peak OS 0 6.84 24.08

FSFC 24.22 25.9 27.05
OFC 22.29 25 22.72
% reduction in ST 7.96 3.47 16.00
3.4 Hydro-thermal interconnected power system
In this section LFC of an interconnected Hydro-thermal power system as shown in Fig. 3.5 is
presented. The procedure described in the section 2.7 is adopted. The simulation parameters are
given in Table 3.7. The power system has 11 state variables. State variables x1 , x5 , x10 and x11
are taken as output feedback states. The optimum gains of OFC and FSFC are obtained. The
computer simulations are carried out with the optimum controller gain settings. MATLAB con-
trol system toolbox [121] is used to simulate the power system and to obtain dynamic responses
of the system for 1% SLP in the control area 1.
TH-1171_07610212
44
0.015
FSFC
OFC
0.01
0.005
−0.005
−0.01
−0.015
−0.02
−0.025
0 5 10 15 20 25 30
time, s
0.01
FSFC
OFC
0.005
−0.005
−0.01
−0.015
−0.02
0 5 10 15 20 25 30
time, s

−3
x 10
4
FSFC
OFC
−2
−4
−6
−8
−10
−12
0 5 10 15 20 25 30
time, s
Figure 3.4: Dynamic responses of TAIPS with reheat and non-reheat turbine
TH-1171_07610212
45
Table. 3.7: Simulation parameters of hydro-thermal interconnected power system
Parameter Value
Prt1 = Prt2 2000 MW
f 60Hz
H1 = H2 5 MW-s/MVA
D1 = D2 0.00833 puMW/Hz
TSG1 0.08s
TT 1 0.3s
KR1 0.5
TR1 10s
TGH2 0.3s
TRS2 5s
TRH2 28.75s
TW 2 0.3s
R1 = R2 2.4 Hz/puMW
TPS1 = TPS2 20s
KPS1 = KPS2 120 s
B1 = B2 0.425
a12 -1
2π T12 0.215

 
−0.0500 6.0000 0 0 0 0 0 0 −6.0000 0 0
−3.3333
 0 3.3333 0 0 0 0 0 0 0 0

 
 
 −5.2083 0 −0.1000 −6.1500 0 0 0 0 0 0 0 
 
 −10.4167 0 0 −12.5000 0 0 0 0 0 0 0 
 
 0 0 0 0 −0.0500 6.0000 0 0 6.0000 0 0 
 
Ã = 
 0 0 0 0 0.9662 −6.6667 6.7362 1.0899 0 0 0 

0 0 0 0 −0.4831 0 −0.0348 −0.5449 0 0 0
 
 
 0 0 0 0 −2.7778 0 0 −3.3333 0 0 0

 
 
 0.2150 0 0 0 −0.2150 0 0 0 0 0 0 
 
 0.8417 0 0 0 0 0 0 0 1.0000 0 0 
0 0 0 0 0.8417 0 0 0 −1.0000 0 0

0 0 0.5000 1.0000 0 0 0 0 0 0 0
B̃ =
0 0 0 0 0 −0.3478 0.1739 1.0000 0 0 0
and the matrix C̃ is defined as

 
1 0 0 0 0 0 0 0 0 0 0
 
0 0 0 0 1 0 0 0 0 0 0
C̃ = 
 

 0 0 0 0 0 0 0 0 0 1 0 
0 0 0 0 0 0 0 0 0 0 1
TH-1171_07610212
46
Control area-1
B1 1/R1 PD1(s)
+ - -
ACE1 + Pg1(s) 1 Pv1 (s) Pm1(s) KPS1 F1(s)
ACE1 Pc1 (s) 1 (1+sKR1TR1)
1/s (1+sTSG1 )
x10 u1 x4 (1+sTT1)
x3 (1+sTR1) x2 + x1
(1+sTPS1)
+ -
Governor
Reheat Turbine Power system
T12
+
P tie,1
!"#
x9
a12 -
a12
T12
+ Pg2(s) Pm2(s) - KPS2
ACE2 ACE2 Pc2 (s) + 1 Pv2 (s) (1-sTw2 ) F2(s)
(1+sTRs2 )
1/s
x11 u2 (1+sTGH2 )
x8 (1+sTRH2 ) x7 (1+0.5sTw2 ) x6 + x5
(1+sTPS2)
+ - -
Hydro turbine Power system
Governor
B2 1/R2 PD2(s)
Control area-2
Figure 3.5: Hydro-thermal interconnected power system
 
0.7084 0 0 0 0 0 0 0 0.8417 0
 0 0 0 0 0 0 0 0 0 0 0

 
 
 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0.7084 0 0 0 −0.8417 0 0 
 
Q̃ = 
 0 0 0 0 0 0 0 0 0 0 0 

0 0 0 0 0 0 0 0 0 0 0
 
 
 0 0 0 0 0 0 0 0 0 0 0

 
 
 0.8417 0 0 0 −0.8417 0 0 0 2.0000 0 0 
 
 0 0 0 0 0 0 0 0 0 1.0000 0 
0 0 0 0 0 0 0 0 0 0 1.0000

1 0
R̃ =
0 1
The optimal gains of FSFC K̄ and OFC K̃ are obtained as:

0.2713 0.5047 7.1418 −3.4754 0.0370 0.0432 1.1489 −0.1458 0.0062 0.9974 0.0721
K̄ =
0.2007 0.2805 1.1920 −0.5571 0.3878 0.4157 11.3685 −1.5619 2.7044 −0.0721 0.9974

0.0462 0.0770 0.7861 0.1915
K̃ =
−0.5856 0.0934 0.2036 0.4272
Dynamic responses of the system are obtained for 1% SLP in the control area 1. The dy-
TH-1171_07610212
47

FSFC -0.02385 -0.03695 -0.004937
OFC -0.02385 -0.02583 -0.003258
% reduction in Peak OS 0 30.09 34.00

FSFC 50 50 50
OFC 16.85 18.91 43.34
% reduction in ST 66.30 62.18 13.32
namic responses are shown in Fig. 3.6. It is observed that the OFC gives better dynamic
responses having relatively smaller peak overshoot and lesser settling time with zero steady
state error as compared to that of FSFC. The quantitative comparison is made in Table 3.8 and
3.9 where percentage improvement in the peak OS of ∆ f2 and ∆Ptie,1 is 30.09 and 34.00, re-
spectively and percentage improvement in the settling time(ST) of ∆ f1 , ∆ f2 and ∆Ptie,1 is 66.3,
62.18, and 13.32, respectively. Moreover the dynamic responses obtained with FSFC are more
oscillatory. The proposed OFC competes well and thus the dynamic responses obtained satisfy
the LFC requirements.
TH-1171_07610212
48
0.04
FSFC
OFC
0.03
0.02
0.01
−0.01
−0.02
−0.03
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
0.06
FSFC
OFC
0.04
0.02
−0.02
−0.04
−0.06
0 5 10 15 20 25 30 35 40 45 50
time, s

−3
x 10
8
FSFC
OFC
−2
−4
−6
−8
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 3.6: Dynamic response of hydro-thermal interconnected power system
TH-1171_07610212
49
3.5 Summary
The LFC of various TAIPSs has been studied using OFC. The combinations of non-reheat,
reheat and hydro turbines are taken to demonstrate the suitability of OFC for LFC of inter-
connected power systems with variety. The Dynamic responses obtained with OFC have been
compared with that of FSFC. The dynamic responses are obtained for thermal-thermal with non-
reheat turbine, thermal-thermal considering reheat turbine and hydro-thermal interconnected
power systems. It has been found that dynamic responses obtained with proposed OFC are
comparatively better and competes well with FSFC, satisfying the LFC requirements. The OFC
is presented from pragmatic point of view which uses less number of states as feedback, thereby
reducing the controller cost and complexities.
TH-1171_07610212
50
C HAPTER 4
LFC OF POWER SYSTEMS WITH
MULTI - SOURCE POWER GENERATION
4.1 Introduction
LFC is an important function in modern Energy management systems. Most of the researchers
considered either thermal or hydro generating units in a control area [1, 9, 13, 15, 124]. In a real
situation, control area may have variety of sources of generations such as hydro, thermal, gas,
nuclear, solar, wind etc. The control area having different sources of power generation repre-
sented by an equivalent of thermal or hydro unit dynamics only may not result in realistic design
of LFC control [6,89,116]. Furthermore, in the new power system environment, generators may
or may not contribute in the LFC task and contribution factors are not the same for all partici-
pant generators [1,5,6,17,89,116]. Keeping in view the present power scenario, combination of
multi-source generators in a control area with their corresponding generation contribution fac-
tors is more realistic for the study of LFC [89, 116]. In a competitive environment, generation
contribution factors are actually time-dependent variables and must be computed dynamically
based on bid prices, availability, congestion problems, costs and other related issues.
Most recently the LFC of power systems with MSPG is reported [89, 116, 117]. Challa et
al. [89] has presented the analysis and design of controller for two area hydro-thermal-gas LFC
system. They have shown that for LFC study, optimal PI state feedback controller is more robust
and performs better than conventional genetic algorithm based PI controller. However, this
optimal PI state feedback controller uses all the states for feedback purpose which is practically
TH-1171_07610212
51
Chapter 4 LFC of power systems with multi-source power generation
1/R11
Reheat-Turbine Thermal Power Plant
B1 -
x3 x2 Pm11(s) PD1(s)
Pc11(s) + 1 x4 (1+sKR1TR1) 1
11
(1+sTSG1) (1+sTR1) (1+sTT1)
1 + -
Pm21(s) KPS1 x1
s 1 x6 x5 + +
Pc21(s) +
x7 (1+sTRS1) (1-sTW1)
21
(1+sTPS1) F1(s)
x12 (1+sTGH1) (1+sTRH1) (1+0.5sTW1)
+
-
Hydro Power Plant with Mechanical Hydraulic Governor
1/R21
Pc31(s) + 1 x11 (1+sXG1) x10 (1-sTCR1) x9 1 x8 Pm31(s)

31
(cg1 +bg1s) (1+sYG1 ) (1+sTF1) (1+sTCD1)
-
Gas Turbine Power Plant
1/R31
Figure 4.1: Block diagram of the single area Power System comprising Reheat-Thermal, Hydro
and Gas generating units
difficult and results in the increased complexity and cost of the controller.
Keeping in view the above, two power system models (i) single area power system with
MSPG and (ii) TAIPS with MSPG are presented for LFC analysis. Each control area of the
power systems includes dynamics of thermal-reheat turbine, hydro and gas based power plants.
The OFC is extended to these systems with pragmatic point of view as it uses less number of
states and results in reduced cost and complexity. An extensive analysis is done considering the
various effects of variation in system parameters and load conditions. To show the effectiveness
of the proposed controller on the actual power system, the LFC of hydro power plants [1] has
also been presented.
4.2 Single area power system with MSPG
The Single area power system (SAPS) with MSPG proposed for study comprises reheat-thermal,
hydro and gas generating units. The linearized models of governors, reheat-turbines, hydro tur-
bines, gas turbines described in [7, 10, 35–37, 116, 125], are taken for simulation of the power
system shown in Fig. 4.1. The simulation parameters and data are given in Table 4.1. The nom-
TH-1171_07610212
52
inal load of the area is taken 1840 MW. The system has n = 12 state variables where x1 = ∆ f1
and x12 = ∫ ACEdt = ∫ B1 ∆ f1 dt. The power system can be described in the state space form
given by the equations (2.52) and (2.38). The associated vectors and matrices are as follows:
state vector, x = [x1 x2 .......x12]T ,
control vector, u = [u1 u2 u3 ]T = [∆Pc11 ∆Pc21 ∆Pc31 ]T
disturbance vector, w = [w1 ] = [∆PD1 ].
KPS1 α11 KPS1 α21 KPS1 α31

 
−T1 TPS1 0 0 TPS1 0 0 TPS1 0 0 0 0
PS1
0 −1 1 0 0 0 0 0 0 0 0 0
 TT 1 TT 1 
−KR1 KR1
 TSG1 R11 0 − T1 1
TR1 − TSG1 0 0 0 0 0 0 0 0 
 R1 
−1 0 0 −T 1 0 0 0 0 0 0 0 0
 TSG1 R11 SG1

2TRS1 2TRS1
−T2 2 2 2
 
 TRH1 TGH1 R21 0 0 0
W1 TW 1 + TRH1 TRH1 TGH1 − TRH1 0 0 0 0 0 
−TRS1 1 1 − TRS1
 TRH1 TGH1 R21 0 0 0 0 −T TRH1 TRH1 TGH1 0 0 0 0 0 
Ã = 
 −1
TGH1 R21 0 0 0 0
RH1
0 −T 1 0 0 0 0 0


GH1
−T 1 1
 0 0 0 0 0 0 0 0 0 0

 CD1 TCD1 
TCR1 XG1 TCR1 TCR1 XG1 cg1 TCR1
 TF1 R31 YG1 bg1 0 0 0 0 0 0 0 − T1 1
TF1 + TF1 YG1 TF1 YG1 bg1 − TF1 YG1 0 
 F1 
−XG1
0 0 0 0 0 0 0 0 −Y 1 1 − XG1 cg1 0
 R31 YG1 bg1 G1 YG1 YG1 bg1

c
− bg1
 −1 0 0 0 0 0 0 0 0 0 0

R31 bg1 g1
B1 0 0 0 0 0 0 0 0 0 0 0
 KR1 1

0 0 TSG1 TSG1 0 0 0 0 0 0 0 0
B̃T =  0 0 0 0 −T 2TRS1
RH1 TGH1
TRS1
TRH1 TGH1
1
TGH1 0 0 0 0 0 
0 0 0 0 0 0 0 0 − TXG1YTCR1 XG1 1
0
F1 G1 bg1 YG1 bg1 bg1
F̃ T =

− KTPS1 0 0 0 0 0 0 0 0 0 0 0
PS1
and the output matrix C̃ is defined as

C̃ = 0 0 0 0 0 0 0 0 0 0 0 1
The performance index J is given in equation(2.39), the matrices Q̃ and R̃ are defined for
this problem using the design criterion given in the section 2.6.2. Excursions of ACE (B1 x1 ),
∫ ACEdt (x12 ) and control vectors(u1 , u2 , u3 ) about their steady values are minimized and J can
be written as:
Z∞ h
1 i
J= (B1 x1 )2 + (x12 )2 + u21 + u22 + u23 dt (4.1)
2
0
TH-1171_07610212
53
From the equation (4.1), the matrices Q̃ and R̃ can be recognized as:
 
B1 2 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
 
 
 0 0 0 0 0 0 0 0 0 0 0 0

 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
Q̃ =  

 0 0 0 0 0 0 0 0 0 0 0 0 

 0 0 0 0 0 0 0 0 0 0 0 0

 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 0 
0 0 0 0 0 0 0 0 0 0 0 1
and
 
1 0 0
R̃ =  0 1 0 
0 0 1
The optimal gain of the proposed OFC is obtained as

 
0.1514
K̃ =  0.0131 
0.0708
Dynamic responses of the system are obtained for 1% SLP in the area through MATLAB sim-
ulation. The frequency deviation response is depicted in Fig. 4.2. It has been observed that the
OFC gives better frequency deviation response having relatively smaller peak overshoot and
lesser settling time with zero steady state error as compared to the FSFC. As given in Table
4.2, percentage improvement in peak OS and settling time with OFC is becoming 6.38 and
61.66, respectively. Some steady state error remains if there is no controller in the supplemen-
tary loop. The output power deviation responses of thermal, hydro and gas units are shown in
Fig. 4.3. The generators respond to accommodate the change in area load by increasing their
corresponding generations. The power generations with OFC in action, attains desirable steady
values to match the local area load change. The effects of generation rate constraints, selection
of regulation parameters and sensitivity analysis are discussed in sections (4.2.2), (4.2.3) and
TH-1171_07610212
54
(4.2.4), respectively.
Table. 4.1: Simulation parameters of SAPS with MSPG
Parameters Value
Prt1 2000 MW
PL1 1840 MW
f1 = f 60 Hz
H1 5 MW-s/MVA
D1 0.0153 pu MW/Hz
KPS1 65.2174 Hz/pu MW
TPS1 10.8696 s
TSG1 0.08 s
TT 1 0.3 s
TCD1 0.2 s
B1 1
KR1 0.3
TR1 10 s
TW 1 1s
TRS1 5s
TRH1 28.75 s
TGH1 0.2 s
XG1 0.6 s
YG1 1s
cg1 1
bg1 0.05 s
TF1 0.23 s
TCR1 0.01s
R1 = R 2.4 Hz/pu MW
α11 0.543478
α21 0.326084
α31 0.130438
4.2.2 Effect of generation rate constraint
In most of the research papers, the effect of restriction on the rate of change of power generation
is not considered [8, 19, 39, 86, 91, 92]. In power systems having steam plants and hydro plants,
power generation can change only at a specified maximum rate [2, 81, 85, 124–126]. Most of
the reheat units have a generation rate around 3 percent/minute. Some have a GRC between 5
and 10 percent/min. For testing further the effectiveness of the proposed controller, the GRC for
thermal and hydro units is taken into account in the computer simulation model. GRC for hydro
unit: for raise 270 percent/minute (0.045pu/s) and for lower 360 percent/minute (0.06pu/s) is
TH-1171_07610212
55
0.02
0.01
Frequency deviation (Hz)

−0.01
−0.02
−0.03
Without supplementary control

−0.04
FSFC
OFC
−0.05
−0.06
−0.07
−0.08
0 10 20 30 40 50 60
time, s
Figure 4.2: Frequency deviation response to 1% SLP in the area
Table. 4.2: Dynamic response comparison of a SAPS
Peak OS of ∆ f1 (Hz) ST of ∆ f1 (s)

FSFC -0.0668 60
OFC -0.06242 23
% reduction 6.38 61.66
considered and GRC for reheat turbine thermal unit: for raise and lower 10 percent/minute
(0.0017pu/s) is considered for study. The proposed power system is simulated with and without
the above GRC limits [9,81,114,123] in thermal and hydro power generating units. The turbine
model considering GRC is shown in Fig. 4.5. An attempt is made to show that how does the
proposed controller designed on linearized models behave if non-linearity is introduced. The
effect of GRC on the frequency deviation response of the area obtained with OFC at nominal
load is shown in the Fig. 4.4. There is a little increase in peak OS, however settling time remains
same. The frequency deviations settles with zero steady state erro. It has been observed that
proposed controller performs well and can accommodate the effect of GRC. Therefore, dynamic
response of the system with GRC satisfies LFC problem requirement.
4.2.3 Effect of governor speed regulation parameter (R)
Fig. 4.6 shows the frequency deviation response to 1% load perturbation in the area at nominal
load with R varying from 1% to 8%. It is assumed that Rki = Ri and Ri = R. It has been observed
TH-1171_07610212
56
−3
x 10
Deviation in thermal power generation(pu MW)

9
6 FSFC
5 OFC
−1
0 10 20 30 40 50 60 70 80
time, s
(a) Thermal power generation deviation response

−3
x 10
4
Deviation in hydro power generation (pu MW)
3.5
3 FSFC
OFC
2.5
1.5
0.5
−0.5
0 10 20 30 40 50 60 70 80
time, s
(b) Hydro power generation deviation response

−3
x 10
4
FSFC
Deviation in gas power generation (pu MW)
3.5 OFC
2.5
1.5
0.5
−0.5
0 10 20 30 40 50 60 70 80
time, s
(c) Gas power generation deviation response
Figure 4.3: Power generation deviation responses to 1% SLP in the area
TH-1171_07610212
57
0.02
Frequency deviation (Hz) −0.02

Without GRC
With GRC
−0.04
−0.06
−0.08
−0.1
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 4.4: Frequency deviation response to 1% SLP in the area considering GRC
+ !m(s)
!v(s)
1 1
TT s
+ -
-
Figure 4.5: Block diagram of a non-linear turbine(GRC)
that for higher values of R i.e. 6% and 8%, the peak overshoot and settling time increases; if
we further increase the value of R, the system gives larger frequency deviation oscillations and
settling time. For lower values of R i.e. 1% and 2% the system becomes more oscillatory and
results in increased cost of governors [15]. Therefore, there should be a compromise between
the governor cost and dynamic performance. Findings reveal that there is no need of going
for the lower values and the higher values of R, since medium value of R with corresponding
optimum controller gains can be preferred to provide better dynamic response. Therefore R=4%
taken for the LFC of the proposed system.
4.2.4 Sensitivity analysis
Lesser attention has been paid to this aspect of LFC problem. Most recently Debbarma et
al. [114] carried out the sensitivity analysis for LFC using BF technique based fractional-order
TH-1171_07610212
58
0.04
0.02
Frequency Deviation (Hz)

0
−0.02
−0.04
R=8%
R=6%
R=4%
−0.06
R=2%
R=1%
−0.08
−0.1
0 2 4 6 8 10 12 14 16 18 20
time, s
Figure 4.6: Frequency deviation response to 1% SLP in the area with variation in R from 1% to
8%
Table. 4.3: Sensitivity Analysis

Parameter % Optimal output feedback Performance
variation change controller gains index J0
K̃11 K̃21 K̃31
All nominal 0 0.1514 0.0131 0.0708 0.414996
Loading +25 0.1537 0.0142 0.0719 0.406347
condition −25 0.1491 0.0119 0.0696 0.424447
TSG +25 0.1496 0.0130 0.0708 0.419149
TSG −25 0.1532 0.0131 0.0708 0.410883
TGH +25 0.1510 0.0126 0.0706 0.416183
TGH −25 0.1519 0.0136 0.0711 0.413668
TR +25 0.1540 0.0151 0.0758 0.41860
TR −25 0.1482 0.0107 0.0643 0.409793
TT +25 0.1454 0.0126 0.0706 0.429905
TT −25 0.1577 0.0135 0.0709 0.400242
TRH +25 0.1515 0.0103 0.0713 0.414982
TRH −25 0.1508 0.0175 0.0698 0.417074
TW +25 0.1503 0.0051 0.0698 0.426905
TW −25 0.1526 0.0221 0.0718 0.402864
TCD +25 0.1521 0.0218 0.0694 0.406546
TCD −25 0.1531 0.0224 0.0743 0.399133
PID controller. Sensitivity analysis is carried out to demonstrate the robustness of the optimum
OFC gains (at nominal condition) for wide changes in system parameters and loading condition
from their nominal values. The system parameters and operating load condition are varied by
±25% from their nominal values, taking one at a time. Table 4.3 gives the optimum controller
gain settings for varied system parameters and operating load condition using optimal OFC.
Dynamic responses are presented in Figs.4.7 and 4.8 for each changed conditions with their
corresponding K̃ and further compared with the responses obtained with K̃ at nominal condition.
Critical examination of all the responses clearly reveals that responses are more and less same.
TH-1171_07610212
59
Effect of Load conditions

0.02
0.01

0
−0.01
−0.02 Nominal load

−0.03 −25% of Nominal load
+25% of Nominal Load
−0.04
−0.05
−0.06
−0.07
0 5 10 15 20 25
time, s
Figure 4.7: Frequency deviation response with variation in load conditions
Effect of speed governor time constant

0.01
0
−0.01
−0.02
−0.03 Nominal Vaue of the Speed Governor time constant

−25%of the Nominal value
−0.04 +25% of the Nominal value
−0.05
−0.06
−0.07
0 5 10 15 20 25
time, s
Figure 4.8: Frequency deviation response with variation in speed governor time constant
Only two number of responses are shown in Figs.4.7 and 4.8 to justify the statement. Thus,
K̃ obtained at nominal parameter and nominal loading condition need not be reset for wide
changes in the system loading or system parameters.
4.3 Two-area interconnected power system with MSPG
Having studied the LFC of SAPS with MSPG, the work has been extended for interconnected
power systems. The TAIPS simulated in this study is shown in Fig. 4.9 which comprises reheat-
thermal, hydro and gas generating units in each control area. The simulation parameters and
data are given in Table 4.4. The nominal load of the control area is taken 1640 MW. The power
system has 25 state variables. State variables x1 , x12 , x24 and x25 are taken as output feedback
TH-1171_07610212
60
Table. 4.4: Simulation parameters of TAIPS with MSPG
Parameters Value
Prt1 = Prt2 2000 MW
PL1 = PL1 1640 MW
f 60 Hz
H1 = H2 5 MW-s/MVA
D1 =D2 0.0137 pu MW/Hz
KPS1 = KPS2 73.1707 Hz/pu MW
T12 0.0433
TPS1 =TPS2 12.1951 s
TSG1 =TSG2 0.08 s
TT 1 =TT 2 0.3 s
TCD1 = TCD2 0.2 s
B1 =B2 0.4303
a12 -1
KR1 = KR2 0.3
TR1 = TR2 10 s
TW 1 = TW 2 1s
TRS1 = TRS2 5s
TRH1 =TRH2 28.75 s
TGH1 = TGH2 0.2 s
XG1 = XG2 0.6 s
YG1 =YG2 1s
cg1 =cg2 1
bg1 = cg2 0.05 s
TF1 = TF2 0.23 s
TCR1 = TCR2 0.01s
R1 = R2 2.4 Hz/pu MW
α11 = α12 0.579268
α21 = α22 0.274390
α31 = α32 0.146342
states. The power system can be described in the state space form given by the equations (2.52)
and (2.38). The associated vectors and matrices are as follows:

state vector, x = [x1 x2 .......x25]T
control vector, u = [u1 u2 u3 u4 u5 u6 ]T =[∆Pc11 ∆Pc21 ∆Pc31 ∆Pc12 ∆Pc22 ∆Pc32 ]T
disturbance vector, w = [w1 w2 ]T =[∆PD1 ∆PD2 ]T .
and the constant matrices Ã, B̃ and C̃, the disturbance matrix F̃ and design matrices Q̃ and R̃ are
given in Appendix A.1.

The regulation parameter R is taken 4% for all and for the TAIPS, ∆Ptie,1 = ∆Ptie,12 . The
optimum gains of FSFC and optimal OFC have been obtained by running the MATLAB codes
TH-1171_07610212
61
generated on the basis of method described in the controller design section 2.6. MATLAB con-
trol system toolbox [121] is used to simulate the power system and to obtain dynamic responses
of the system for 1% SLPs in the control area-1. The MATLAB Simulations are run with the
optimum controller gains.
1/R11
Control area-1
Reheat-Turbine Thermal Power Plant !D1(s)
B1 -
!c11(s) !m11(s)
+ 1 x4 (1+sKR1TR1) x3 1 x2
(1+sTSG1) (1+sTR1) (1+sTT1) 11
+
+
-
!m21(s) + KPS1 x1
x24 + 1 x7 (1+sTRS1) x6 (1-sTW1) x5 +
1
(1+sTGH1) (1+sTRH1) (1+0.5sTW1) 21 (1+sTPS1) !1(s)
s !c21(s)
- + -
+ Hydro Power Plant with Mechanical Hydraulic Governor
1/R21
!c31(s) x11 x10 x9 x8 !m31(s)

+ 1 (1+sXG1) (1-sTCR1) 1
(cg1+ bg1.s) (1+sYG1) (1+sTF1) (1+sTCD1) 31
-
1/R31
x23
+
!tie, 1(s) 2. !"12
s
-
1/R12
a12 Reheat-Turbine Thermal Power Plant
- a12
!c12(s) + 1 x15 (1+sKR2TR2) x14 x13 !m12(s)
1
(1+sTSG2) (1+sTR2) (1+sTT2) 12
+
!m22(s) + - x12
x18 x17 + KPS2
x25 + 1 (1+sTRS2) (1-sTW2) x16
1 (1+sTGH2) (1+sTRH2) (1+0.5sTW2) 22 (1+sTPS2) !2(s)
s !c22(s) +
+
+ - Hydro Power Plant with Mechanical Hydraulic Governor
-
!D2(s)
1/R22
B2 !c32(s) x22 x21

+ 1 (1+sXG2) (1-sTCR2) x20 1 x19
(cg2 + bg2.s) (1+sYG2) (1+sTF2) (1+sTCD2) 32
!m32(s)
-
1/R32
Control area-2
Figure 4.9: TAIPS comprising reheat-thermal, hydro and gas generating units in each control
area
TH-1171_07610212
62
The optimum value of the K̃ for the proposed OFC corresponding to nominal system parameters
is obtained as:
 
 0.4202 −0.0936 0.3233 0.0808 
 

 −0.1634 −0.0039 −0.0386 −0.0522 

 
0.1622 −0.0597 0.1340 0.0207 
 

K̃ = 




 0.0719 0.8219 −0.1455 0.3840 
 

 −0.0469 −0.6530 −0.1181 −0.4162 

 
0.0473 0.3588 −0.0074 0.2008
Dynamic responses of the system are obtained for 1% SLP in the control area-1 through com-
puter simulation. The frequency deviation and tie line power deviation responses are shown in
Fig. 4.10. It is observed that the OFC gives better dynamic response having relatively smaller
peak overshoot and lesser settling time with zero steady state error as compared to the FSFC.
The dynamic response of the power system considering low, medium and high head hydro tur-
bines is as shown in Fig. 4.11. The results show that frequency and tie line power deviations
are more and less same for low and high head hydro turbines. The settling time increases in
the case of medium head turbine. However OFC shows its suitability for low, medium and high
head hydro plants and gives satisfactory dynamic responses.
4.3.2 Effect of GRC
In this section, the performance of the proposed OFC is evaluated considering GRC in power
system simulation model. GRC for Hydro unit: for raise 0.045 pu/s and for lower 0.06 pu/s is
considered and GRC for reheat turbine thermal unit: 0.0017 pu/s is considered for study. The
power system is simulated with and without the above GRC limits in thermal and hydro power
generating units in each area. The effect of GRC on the dynamic responses is shown in the Fig.
4.12. The critical examination of dynamic responses reveals that GRC results in increased peak
overshoot, however settling time remains same for this particular power system model. The
proposed controller accommodates the effect of GRC and gives dynamic responses satisfying
TH-1171_07610212
63
0.005
Frequency Deviation,area−1(Hz)
0
−0.005 Output feedback

Full state feedback
−0.01
−0.015
−0.02
−0.025
−0.03
−0.035
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
(a) Frequency deviation response of area-1
0.005
0
−0.005
Output feedback
Full state feedback
−0.01
−0.015
−0.02
−0.025
−0.03
−0.035
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
(b) Frequency deviation response of area-2
−3
x 10
4
Output feedback
Full state feedback
Tie line Power Deviation (pu MW)
−2
−4
−6
−8
0 10 20 30 40 50 60
time, s
Figure 4.10: Comparison of frequency and tie line power deviation responses for 1% SLP
TH-1171_07610212
64
0.005
Frequency deviation, area−1(Hz)

−0.005
Low head
Medium head
−0.01
High head
−0.015
−0.02
−0.025
−0.03
−0.035
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
0.005
−0.005
Low head
Medium head
−0.01 High head
−0.015
−0.02
−0.025
−0.03
0 5 10 15 20 25 30 35 40 45 50
time, s

−3
x 10
3
Low head
2 Medium head
High head
1
−1
−2
−3
−4
−5
−6
−7
−8
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 4.11: Frequency and tie line power deviation responses with low, medium and high head
hydro turbines
TH-1171_07610212
65

0
With GRC
−0.01 Without GRC
−0.02
−0.03
−0.04
−0.05
0 5 10 15 20 25 30 35 40 45 50
time, s

With GRC
−0.01
Without GRC
−0.02
−0.03
−0.04
−0.05
−0.06
0 5 10 15 20 25 30 35 40 45 50
time, s

−3
x 10
5
With GRC
Without GRC
−5
−10
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 4.12: Dynamic responses considering the GRC
TH-1171_07610212
66
LFC problem requirements. It shows that optimal OFC designed for linearized models, may
perform well even if GRCs are considered.
4.3.3 Selection of governor speed regulation parameter(R)
Fig. 4.13 shows the dynamic response to 1% SLP in the control area-1 with R varying from
2% to 6%. It is assumed that Rki = Ri and for equal area Ri = R . It is observed that for
higher values of R i.e. 5% and 6%, the peak overshoot and settling time increases; if we further
increase the value of R, the system gives larger peak overshoots and settles with larger settling
time. The dynamic response for 3% to 4% value of R results with smaller peak overshoot and
lesser settling time. For further lower values of R the system dynamic response becomes more
oscillatory. Findings reveal that there is no need of going for the lower values and the higher
values of R, since medium value of R with corresponding optimum controller gains can be
preferred to provide better dynamic response of LFC. For this power system R is chosen 4%
which yields improved results.
4.3.4 Effect of variation in load disturbance
In fact, LFC deals with small level disturbances in the power system. In most of the studies,
1% SLP is taken for the study of LFC problem. In real situations, the load disturbances also
vary. In this case, system dynamic responses are obtained for the wide range of SLP varying
from 1% to 5% in control area 1. Dynamic responses for different SLPs are shown in Fig. 4.14.
It is concluded that for SLP varying from 1% to 5%, the first peak overshoot increases with
increase in the level of SLP and settling time remains approximately same with zero steady state
error. The proposed OFC acts fast to settle the deviations in the frequency and tie line power.
Therefore, the proposed controller satisfy the LFC problem requirement for wide variation of
load disturbance.
TH-1171_07610212
67
0.005
0
−0.005
R=2%
R=3%
−0.01
R=4%
R=5%
−0.015 R=6%
−0.02
−0.025
−0.03
−0.035
0 5 10 15 20 25 30 35 40
time, s
0.005
R=2%
−0.005 R=3%
R=4%
R=5%
−0.01
R=6%
−0.015
−0.02
−0.025
−0.03
0 5 10 15 20 25 30 35 40
time, s

−3
x 10
3
2
Tie line Power deviation (pu MW)
−1
R=2%
−2 R=3%
R=4%
−3 R=5%
−4 R=6%
−5
−6
−7
−8
0 5 10 15 20 25 30 35 40
time, s
Figure 4.13: Effect of the R on dynamic responses
TH-1171_07610212
68
0.02

0
−0.02 1% SLP
2% SLP
−0.04
3% SLP
−0.06 4% SLP
5% SLP
−0.08
−0.1
−0.12
−0.14
−0.16
−0.18
0 5 10 15 20 25 30 35 40 45 50
time, s
0.02
−0.02
1% SLP
−0.04 2% SLP
3% SLP
−0.06 4% SLP
5% SLP
−0.08
−0.1
−0.12
−0.14
0 5 10 15 20 25 30 35 40 45 50
time, s
0.01
0.005
−0.005 1% SLP
2% SLP
−0.01
3% SLP
−0.015 4% SLP
5% SLP
−0.02
−0.025
−0.03
−0.035
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 4.14: Frequency and tie line power deviation responses for SLPs in area-1, varying from
1% to 5%
TH-1171_07610212
69
−3
x 10
4
0.005 0.005
Nominal value

0
0 2 +25% of nominal value
−25% of nominal value
−0.005
Nominal value
−0.005 Nominal value
+25% of nominal value 0
−0.01 +25% of nominal value
−0.015 −0.01
−2
−0.02 −0.015
−0.025 −4
−0.02
−0.03
−6
−0.025
−0.035
−0.04 −0.03 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
time, s time, s time, s
(a) Frequency deviation response of area-1 (b) Frequency deviation response of area-2 (c) Tie line power deviation response
Figure 4.15: Effect of variation in steam turbine speed governor time constant(TSG ) on dynamic
responses
−3 −3
x 10 x 10
4
0.005 5 Nominal value
+25% of nominal value

0 −25% of nominal value

2
0
−0.005
Nominal value
0
+25% of nominal value −5
−25% of nominal value −2
−0.015 −10
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.16: Effect of variation in steam turbine time constant(TT ) on dynamic responses
4.3.5 Sensitivity analysis
The system parameters and operating load condition are varied by ±25% from their nominal
values, taking one at a time. Dynamic responses for variation in system parameters and load
condition are shown in Figs.(4.15)-(4.22). An extensive analysis is done to examine the effect
of variation in system parameters and load conditions. Investigations reveal that the dynamic
responses hardly change when system parameters and operating load condition are changed by
±25% from their nominal values with the optimum controller gains once set. It is evident that
OFC accommodates for the wide range of variations in system parameters and load condition
from their nominal values. The settling time and peak overshoot of the dynamic responses under
these varied conditions are more and less same with zero steady state error.
TH-1171_07610212
70
−3 −3
x 10 x 10
4

2
0
−0.005
Nominal value
0
+25% of nominal value −5 Nominal value
−0.01
−25% of nominal value +25% of nominal value
−0.015 −10 −2
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.17: Effect of variation in steam turbine reheat time constant(TR) on dynamic responses
−3 −3
x 10 x 10
4


2
0
−0.005
Nominal value
0
−0.01
−0.015 −10 −2
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.18: Effect of variation in hydro turbine speed governor main servo time constant(TGH )
on dynamic responses
−3 −3
x 10 x 10
4

2
0
−0.005
Nominal value
0
−0.01
−0.015 −10 −2
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.19: Effect of variation in hydro turbine speed governor transient droop time
constant(TRH ) on dynamic responses
TH-1171_07610212
71
−3 −3
x 10 x 10
4

2
0
−0.005
Nominal value
0
−0.01
−0.015 −10 −2
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.20: Effect of variation in gas turbine compressor discharge volume time constant (TCD )
on dynamic responses
−3 −3
x 10 x 10
4


2
0
−0.005
Nominal value
0
+25% of nominal value −5
−25% of nominal value −2
−0.015 −10
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.21: Effect of variation in nominal starting time of water in penstock (TW ) on dynamic
responses
−3 −3
x 10 x 10
4
0.005 5
Nominal value

0
0
−0.005
Nominal value 0
−5
−0.01 +25% of nominal value Nominal value
−0.015 −10 −25% of nominal value −2
−0.02
−15 −4
−0.025
−20 −6
−0.03
−0.035 −25 −8
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Figure 4.22: Effect of variation in nominal area load on dynamic responses
TH-1171_07610212
72
4.4 LFC of hydro power plants
To ensure the good performance of the proposed optimal OFC on an actual power system,
hydro power plants operational in KHOZESTAN (a province in southwest of Iran) [1] are taken
as an additional case study in this section. Dez output power is 520 MW whereas Karoon3
output power is 1000 MW operational and 2000 MW nominal. The control area is linked to
other control areas of an interconnected power system. The control area includes Karoon3 and
Dez hydro power plants where variation of area load and interaction of other control areas are
considered as two disturbances in control area.

Most recently (2012), Alireza et al. [1] studied the LFC of this operational plant using a
robust optimal MISO PID controller where the tuning of PID controller is stated as an opti-
mization problem in which a combination of quadratic index and maximal complex/real ratio of
the closed loop poles is minimized subject to some constraints on characteristic matrix Eigen-
values.
For analysis, their system dynamics, system parameter values, load disturbance conditions
and assumptions are considered here. For clarity and sake of comparison, nomenclature of
system parameters in this section are kept same as in [1]. Block diagram of a control area
including two hydro power plants [1] is referred for formulating the state equations. The block
diagram of the interconnected power system is reproduced [1] for ready reference and shown
in the Fig. 4.24. Refering equations (2.37) and (2.38), in this case study n=7, m=1 and p=3,
where the state vector is x = [∆ f ∆Ptie ∆PtK ∆PtD ∆PgK ∆PgD ∫ ACE]T and the output state vector
is y = [∆ f ∆Ptie ∫ ACE]T .
The system parameters [1] taken to run the MATLAB Simulation are TtK = 0.32s, TtD =
0.24s, D = 0.0534puMW /Hz, RD = RK = 0.467Hz/puMW , β = 4.334puMW /Hz and TgK =
TgD = 11s. System dynamic responses are obtained with optimal OFC gains for two different
situations based on the two types of the present disturbances in the system. In the first situation,
the effects of load disturbances in the area itself are examined, w1 is taken as shown in Fig.
4.24(a) and w2 = 0, whereas in the second situation, the effects of disturbances due to neighbor-
ing interconnected control areas are examined which affect the area under control, w2 is taken
TH-1171_07610212
73
Droop
1
R
- 1 Pg 1 Pt
u +
(1+sTg ) (1+sTt )
Governor Rate limiter Turbine
(a) Block diagram of a Genco

Genco 1, Karoon 3
w1
PtK
Droop
+ ACE
u -
+ 1 f
Controller
+ (M.s+D)
+ -
PtD
Power system
Ts
Droop
Genco 2, Dez
Ptie
1 !
s
-
Frequency bias constant

w2
(b) Block diagram of a control area including two hydro power plants
Figure 4.23: Model of hydro power plants operational in KHOZESTAN (Iran), [1]
Governor
0.1
0.08
MW)
W1 (pu M
0.06
0.04
0.02
0 5 10 15 20 25 30 35 40 45 50
Time(s)
()
(a)
0.025
0.02
MW)
W2 (pu M
0.015
0.01
0.005
0 5 10 15 20 25 30 35 40 45 50
Time(s)
(b)
Figure 4.24: (a)Area load disturbance w1 , first situation (b) Disturbance due to neighboring
interconnected control areas w2 , second situation [1]
TH-1171_07610212
74
as shown in Fig.4.24(b) and w1 = 0. The load disturbance patterns taken in both the situations
are reproduced from [1].

Fig. 4.25 shows the dynamic responses obtained using proposed OFC controller. As com-
pared, the trajectories of deviation in frequency and tie line power are more and less same. Peak
overshoots in frequency deviations observed at each time instant of load changes are improved
remarkably and reduction is becoming more than 80%. The improvement in settling time is
also observed.
Like MISO PID controller, the proposed controller also guarantees the good performance,
such as frequency deviation elimination and disturbance attenuation as well as robustness under
area load changes or frequency variation of interconnected areas scenarios. As compared, the
peak overshoots in frequency deviation responses obtained with proposed controller are very
less resulting in remarkable improvement in the frequency deviation response which is the most
important parameter. It is observed that frequency and tie line power settles very fast with zero
steady state error. Therefore, the proposed OFC controller satisfies the LFC requirements.
TH-1171_07610212
75
−3
x 10
6
Frequency Deviation(Hz)
4
−2
−4
−6
0 5 10 15 20 25 30 35 40 45 50
time, s
(a) Frequency deviation response for first situation

Frequency Deviation(Hz)
−0.005
−0.01
−0.015
−0.02
−0.025
−0.03
0 5 10 15 20 25 30 35 40 45 50
time, s
(b) Frequency deviation response for second situation

Tie line power Deviation (pu MW)
0.02
0.01
−0.01
−0.02
0 5 10 15 20 25 30 35 40 45 50
time, s
(c) Tie line Power deviation response for the first situation
0.2
Tie line power Deviation (pu MW)
0.15
0.1
0.05
0 5 10 15 20 25 30 35 40 45 50
time, s
(d) Tie line Power deviation response for the second situation
0.03
Area control error (pu MW)
0.02
0.01
−0.01
−0.02
−0.03
0 5 10 15 20 25 30 35 40 45 50
time, s
(e) Area control error for the first situation

0.04
Area control error (pu MW)
0.02
−0.02
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
(f) Area control error for the second situation
Figure 4.25: Dynamic responses of hydro power plants operational in KHOZESTAN (Iran)
TH-1171_07610212
76
4.5 Summary
The performance of the proposed controller is demonstrated on the multi-source power systems
and its dynamic responses are compared with FSFC. The OFC gives better dynamic response
having relatively smaller peak overshoot and lesser settling time with zero steady state error as
compared to that of FSFC. The effect of GRC on dynamic response is discussed. The critical ex-
amination of dynamic responses reveals that GRC results in increased peak overshoot, however
settling time remains same for this particular power system model. Investigations reveal that it
is better to prefer the medium value of R i.e. 4% with corresponding optimum controller gains
to provide better dynamic response of LFC for the proposed system. The dynamic responses are
obtained for 1% to 5% SLPs. It is observed that for SLP varying from 1% to 5%, the first peak
overshoot increases with increase in the level of SLP and settling time remains approximately
same with zero steady state error. The sensitivity analysis reveals that the optimum values of
controller gains obtained for nominal system parameters and load condition are quite insensi-
tive to wide parameter variation ±25%. Hence for all practical purposes, the controller is quite
robust. The LFC of hydro power plants operational in Iran has also been studied. The proposed
controller performs well on this system and improves the frequency deviation responses remark-
ably. Like MISO PID controller, the proposed controller also guarantees the good performance,
such as frequency deviation elimination and disturbance attenuation as well as robustness under
area load changes or frequency variation of interconnected areas scenarios. The simulation re-
sults show that proposed control strategy is very effective and guarantees good performance. In
fact, this method provides a control system that satisfies the load frequency control requirements
with a credible dynamic response.
TH-1171_07610212
77
C HAPTER 5
LFC OF POWER SYSTEMS CONSIDERING
AC-DC TIE LINES AND TCPS
5.1 Introduction
Electric power systems are interconnected to make the systems more reliable and robust. In a
multi-area system, power generations and loads are coordinated with each other through the tie
lines among the control areas [2, 10]. An electrical power system consists of many generating
units and many loads while its total load demand varies continuously throughout the day. LFC is
one of the important control problems in the interconnected power system design and operation,
and is becoming more significant now due to the increasing size, changing structure, emerging
multi-energy sources and new uncertainties, challenges, environmental constraints, and com-
plexity of power systems [3, 6]. Literature survey shows that mostly AC tie lines are used for
the interconnection of MAIPSs and lesser attention is given to AC-DC parallel tie lines [2, 32].
HVDC transmission system is finding more space in the transmission system due to its various
techno-economical advantages. One of the major applications of HVDC transmission is operat-
ing a DC link in parallel with an AC link interconnecting two control areas to get an improved
system dynamic performance with greater stability margins under small perturbations in the
system [102, 103, 127, 128]. Some research work is reported on LFC of interconnected power
systems connected via HVDC link in parallel with AC link [4, 127]. The idea of LFC of an in-
terconnected power system with a DC tie line in parallel with an AC tie line [77, 102, 127, 128]
is extended for the study of LFC of the TAIPS with MSPG as shown in Figs. 5.1 and 5.2.
TH-1171_07610212
78
Chapter 5 LFC of power systems considering AC-DC tie lines and TCPS
The Flexible AC Transmission Systems (FACTS) devices provide more flexibility in power
system operation and control. Thyristor controlled phase shifter (TCPS) is an effective FACTS
device for the tie line power flow control of an interconnected power system. The TCPS device
is modeled and used in series with tie lines to improve the dynamic performance of LFC of the
interconnected power systems [106, 108].
Literature survey shows that most of the researchers applied optimal control theory on
thermal-thermal power systems with AC tie lines only [8, 39, 49, 81, 114, 124]. Optimal con-
trollers have been used in LFC as secondary controllers, but there is hardly any literature that
compares performances of optimal controllers for the power system with and without DC tie
lines [4]. Some researchers have studied the LFC of thermal-thermal or hydro-thermal power
systems considering TCPS but there is hardly any research work that applies optimal OFC strat-
egy for the LFC of the interconnected power systems with MSPG considering TCPS [4,93,106].
Therefore, there is a scope to study the LFC with MSPG and TCPS.
This chapter presents the LFC study of the interconnected power system considering AC-DC
parallel tie line and TCPS. An attempt is made to improve the dynamical response of the LFC
problem from a practical point of view by considering the OFC strategy using AC-DC parallel
tie lines and TCPS. An extensive analysis is done to study the LFC performance of the OFC for
the interconnected power system models in new power system environment considering AC-DC
parallel tie lines and TCPS. In the section 5.2, a TAIPS model comprising hydro, thermal with
reheat-turbine and gas units in each control area with AC-DC parallel tie line as shown in Fig.
5.2 is presented. HVDC link is connected in parallel with the existing AC link for stabilizing the
frequency oscillations of AC system. Further in the section 5.3.2, a TAIPS model comprising
hydro, thermal with reheat-turbine and gas units in each control area as shown in Fig. 5.7 is
presented. TCPS is connected in series with the AC tie line for stabilizing the control area
frequency and tie line power deviations [93, 106]. The power system simulation is done using
MATLAB simulink and control problem is solved using MATLAB programming. The effect of
the load disturbance by varying the SLP over a wide range from 1% to 4% is examined.
TH-1171_07610212
79
AC Tie line DC Link
Control area-1 Control area-2

Thermal, Hydro and Thermal, Hydro and
Gas Gas
Converter Converter
(Area-1) (Area-2)
Figure 5.1: TAIPS with AC-DC parallel tie lines
5.2 Simulation of the power system with parallel AC-DC tie
lines
The schematic of TAIPS model with AC-DC tie lines is shown in Fig. 5.1. The TAIPS with
MSPG considering parallel AC-DC tie lines as shown in Fig. 5.2 is proposed for LFC study in
this section. As shown in Fig. 5.2, each control area comprises reheat-thermal, hydro and gas
generating units and the two equal areas are interconnected by AC-DC tie lines. The simula-
tion of this interconnected power system in a new power system environment is based on the
concepts of considering variety of generators with their corresponding generation contribution
factors in each control area [6, 89, 117] and parallel AC-DC tie lines [77, 102]. The DC link is
assumed bidirectional as most of the converters used for HVDC are intrinsically able to operate
with power conversion in either direction. The system parameter values are given in Table 5.1.
It is assumed that Rki = Ri . For the TAIPS, ∆Ptie,1 = ∆Ptie,12 . The nominal loading of each con-
trol area is taken 1640 MW with the power generation scheduling and generation contribution
factors as given in Table 5.2.
The state space form is obtained for the proposed power system model. The power system
has 27 state variables. Incremental DC power flow is considered as an additional state variable
in the LFC strategy. State variables x1 , x12 , x26 and x27 are taken as output feedback states. The
state equations can be organized in the state-space form as described by the equations (2.52)
and (2.38).
where, the associated vectors and matrices are as follows:

TH-1171_07610212
80
1/R11
Control area-1 HVDC Link
Reheat-Turbine Thermal Power Plant !DC1(s) KDC1
B1 - !D1(s)
!c11(s) !m11(s) (1+sTDC1)
+ 1 x4 (1+sKR1TR1) x3 1 x2 x24
(1+sTSG1) (1+sTR1) (1+sTT1) 11
+ - -
+ x1
x26 !m21(s) + + KPS1
+ 1 x7 (1+sTRS1) x6 (1-sTW1) x5
1
s !c21(s) + -
1/R21
!c31(s) x11 x10 x9 x8 !m31(s)

+ 1 (1+sXG1) (1-sTCR1) 1
(cg1+ bg1.s) (1+sYG1) (1+sTF1) (1+sTCD1) 31
-
1/R31
x23
+
!tie, 1(s) 2. !"12
s
-
1/R12
- a12
!c12(s) + 1 x15 (1+sKR2TR2) x14 x13 !m12(s)
1
(1+sTSG2) (1+sTR2) (1+sTT2) 12
+
!m22(s) + - x12
x18 x17 + KPS2
x27 + 1 (1+sTRS2) (1-sTW2) x16
s !c22(s) + -
+
+ - -
Hydro Power Plant with Mechanical Hydraulic Governor !D2(s)
1/R22
B2 !c32(s) x22 x21

+ 1 (1+sXG2) (1-sTCR2) x20 1 x19 !DC2(s) KDC2
!m32(s) x25 (1+sTDC2)
-
Gas Turbine Power Plant HVDC Link
1/R32
Control area-2
Figure 5.2: Block diagram of a TAIPS with MSPG considering AC-DC parallel tie lines
TH-1171_07610212
81
Table. 5.1: Simulation Parameters of TAIPS with MSPG considering parallel AC-DC tie lines
Parameters Value
Prt1 = Prt2 2000 MW
PL1 = PL1 1640 MW
f 60 Hz
H1 = H2 5 MW-s/MVA
D1 =D2 0.0137 pu MW/Hz
T12 0.0433
TPS1 =TPS2 12.1951 s
TSG1 =TSG2 0.08 s
TT 1 =TT 2 0.3 s
TCD1 = TCD2 0.2 s
B1 =B2 0.4303
a12 -1
KR1 = KR2 0.3
TR1 = TR2 10 s
TW 1 = TW 2 1s
TRS1 = TRS2 5s
TRH1 =TRH2 28.75 s
TGH1 = TGH2 0.2 s
XG1 = XG2 0.6 s
YG1 =YG2 1s
cg1 =cg2 1
bg1 = cg1 0.05 s
TF1 = TF2 0.23 s
TCR1 = TCR2 0.01s
α11 = α12 0.579268
α21 = α22 0.274390
α31 = α32 0.146342
DC link parameters
KDC1 =KDC2 1
TDC1 =TDC2 0.2 s

The optimum gains of optimal OFC are obtained by running the MATLAB codes generated
on the basis of method described in section 2.6. The computer simulations are carried out
TH-1171_07610212
82
with the optimum controller gain settings. MATLAB control system toolbox [121] is used to
simulate the power system and to obtain dynamic responses of the system for 1% SLP in the
control area-1.
The optimum values of the K̃ for the OFC by minimizing the cost function for the power system
with AC tie line corresponding to nominal system parameters is

 
0.4202 −0.0936 0.3233 0.0808
−0.1634 −0.0039 −0.0386 −0.0522
 
 
 
0.1622 −0.0597 0.1340 0.0207
K̃ = 
 

 0.0719 0.8219 −0.1455 0.3840 
 
 −0.0469 −0.6530 −0.1181 −0.4162 
0.0473 0.3588 −0.0074 0.2008
and for the power system with AC-DC parallel tie line is
 
−0.1887 0.5428 0.5702 0.2080
−0.0644 0.1023 0.0895 0.0568
 
 
 
−0.1082 0.2338 0.2685 0.0629
K̃ = 
 

 0.4622 −0.0495 0.2321 0.3059 
 
 −0.2360 0.1406 −0.0580 −0.3128 
0.2028 −0.0828 0.0784 0.1894
Dynamic responses of the system are obtained for 1% SLP in the control area-1. The fre-
quency deviation responses of control area-1 and control area-2 are shown in Figs. 5.3 and 5.4,
respectively. The tie line power deviation response is shown in Fig. 5.5. It is observed that
the OFC considering parallel AC-DC tie lines in power system gives better dynamic responses
having relatively smaller peak overshoot and lesser settling time with zero steady state error as
compared to the power system with AC tie lines only. The quantitative comparison is made in
Table 5.3 where percentage reduction in the peak OS of ∆ f1 , ∆ f2 and ∆Ptie,1 is 60.49, 90 and
44.92, respectively with DC link. Also, system dynamic responses are obtained for the wide
range of SLP varying from 1% to 4% in either control area. Further, frequency deviation and
tie line power responses for different SLPs are shown in Fig. 5.6. It is apparent that for SLP
varying from 1% to 4%, the first peak overshoot increases with increase in the level of SLP
and settling time remains approximately same with zero steady state error. Thus, the proposed
TH-1171_07610212
83
0.005
0
−0.005
AC Tie line
−0.01 AC−DC Tie line
−0.015
−0.02
−0.025
−0.03
−0.035
−0.04
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.3: Frequency deviation response of control area-1 with AC tie line and AC-DC parallel
tie line
controller satisfy the LFC problem requirement for wide variation of load disturbance.
Table. 5.2: Power Generation Scheduling to match the nominal load of the individual control
area
Total control area Thermal Gas Hydro

load contribution contribution contribution
(MW) (MW) (MW) (MW)
1640 950 240 450
Generation contribution factor 0.579268 0.146342 0.274390
Table. 5.3: Dynamic response comparison

With AC tie line -0.0324 -0.0230 -0.0069
With AC-DC tie line -0.0128 -0.0023 -0.0038
% reduction in Peak OS 60.49 90 44.92
5.3 Incremental tie line power flow considering TCPS
TCPS is a device that changes the relative phase angle between the system voltages. The tie
line power flow can be regulated by controlling the phase angle (φ ) to damp out the control area
frequency deviations and improve power system stability [94, 108, 109]. The schematic of the
TAIPS considering a TCPS in series with the tie line is shown in Fig. 5.8.
TH-1171_07610212
84
0.005
Frequency Deviation,area−2(Hz) 0
AC Tie line
−0.005 AC−DC Tie line
−0.01
−0.015
−0.02
−0.025
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.4: Frequency deviation response of control area-2 with AC tie line and AC-DC parallel
tie line
−3
x 10
6
4 AC Tie line
AC−DC Tie line
−2
−4
−6
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.5: Tie line power deviation response with AC tie line and AC-DC parallel tie line
TH-1171_07610212
85
0.01
0
−0.01
1% SLP
−0.02 2% SLP
3% SLP
4% SLP
−0.03
−0.04
−0.05
−0.06
0 10 20 30 40 50 60
time, s
(a) Frequency deviation response of control area-1

−3
x 10
4
−2 1% SLP
2% SLP
−4 3% SLP
4% SLP
−6
−8
−10
−12
0 10 20 30 40 50 60
time, s
(b) Frequency deviation response of control area-2

−3
x 10
2
−2
−4 1% SLP
2% SLP
−6 3% SLP
4% SLP
−8
−10
−12
−14
−16
−18
−20
0 10 20 30 40 50 60
time, s
Figure 5.6: Frequency deviation and tie line power responses for SLP varying from 1% to 4%
TH-1171_07610212
86
1/R11
Control area-1
Reheat-Turbine Thermal Power Plant !D1(s)
B1 -
!c11(s) !m11(s)
+ 1 x4 (1+sKR1TR1) x3 1 x2
(1+sTSG1) (1+sTR1) (1+sTT1) 11
+
+
-
!m21(s) + KPS1 x1
x25 + 1 x7 (1+sTRS1) x6 (1-sTW1) x5 +
1
s !c21(s) + -
1/R21
!c31(s) x11 x10 x9 x8 !m31(s) TCPS

+ 1 (1+sXG1) (1-sTCR1) 1
(cg1+ bg1.s) (1+sYG1) (1+sTF1) (1+sTCD1) 31 x24 T12 . K
- (1+sT )
1/R31
+
+ x23 !"! T
12
P (s) + s
tie,1
-
1/R12
- a12
!c12(s) + 1 x15 (1+sKR2TR2) x14 x13 !m12(s)
1
(1+sTSG2) (1+sTR2) (1+sTT2) 12
+
!m22(s) + - x12
x18 x17 + KPS2
x26 + 1 (1+sTRS2) (1-sTW2) x16
s !c22(s) + + -
+ - -
Hydro Power Plant with Mechanical Hydraulic Governor !D2(s)
1/R22
B2 !c32(s) x22 x21

+ 1 (1+sXG2) (1-sTCR2) x20 1 x19
!m32(s)
-
1/R32
Control area-2
Figure 5.7: TAIPS with MSPG considering TCPS
1’ 1:1! 1 jX tie,12 2
Control area-1 i12 Control area-2
Thermal, Hydro Thermal, Hydro
and Gas and Gas
TCPS Tie line
Va1 ! a1 Va1 "( a1 #!) Va 2 ! a2
Figure 5.8: Schematic of TAIPS considering TCPS in series with tie line
TH-1171_07610212
87
In a conventional interconnected power system, the incremental tie line power flow ∆Ptie,12
from control area-1 to control area-2 can be expressed as [2]

2π T12
∆Ptie,12 (s) = (∆F1 (s) − ∆F2 (s)) (5.1)
s
When a TCPS is connected in series with the tie line as shown in Fig. 5.8, the current flowing
from control area 1 to control area 2 can be expressed [109] as
|Va1 | ∠(δa1 + φ ) − |Va2 | ∠δa2
i12 = (5.2)
jXtie,12
Now, the tie line power becomes
P̃tie,12 − jQ̃tie,12 = [|Va1 | ∠ − (δa1 + φ )] i12 (5.3)
Using equations (5.2) and (5.3) and separating the real and imaginary parts, P̃tie,12 becomes
|Va1 | |Va2 |
P̃tie,12 = sin(δa1 − δa2 + φ ) (5.4)
Xtie,12
Perturbing δa1 , δa2 and φ from their nominal values δa1

0 , δ 0 and φ 0 , respectively,
a2
|Va1 | |Va2 |
∆P̃tie,12 = cos(δa1
0
− δa2
0
+ φ 0 ) sin(∆δa1 − ∆δa2 + ∆φ ) (5.5)
Xtie,12
Since (∆δa1 − ∆δa2 + ∆φ ) is very small, therefore
sin(∆δa1 − ∆δa2 + ∆φ ) ≈ (∆δa1 − ∆δa2 + ∆φ )
|Va1 | |Va2 |
∆P̃tie,12 = cos(δa1
0
− δa2
0
+ φ 0 )(∆δa1 − ∆δa2 + ∆φ ) (5.6)
Xtie,12
Let T̃12 be the synchronizing coefficient with TCPS:

|Va1 | |Va2 |
T̃12 = cos(δa1
0
− δa2
0
+ φ 0) (5.7)
Xtie,12
∆P̃tie,12 can be given as
∆P̃tie,12 = T̃12 (∆δa1 − ∆δa2 ) + T̃12 ∆φ (5.8)
where,
∆δa1 = 2π 0 ∆ f 1 dt and ∆δa2 = 2π 0 ∆ f 2 dt
Rt Rt
Taking the Laplace transform of equation (5.8)

2π T̃12
∆P̃tie,12 (s) = [∆F1 (s) − ∆F2(s)] + T̃12 ∆φ (s) (5.9)
s
TH-1171_07610212
88
As given in equation (5.9) tie line power flow can be controlled by controlling the phase shifter
angle ∆φ . The phase shifter angle ∆φ can be given [107, 108] as

KΦ
∆φ (s) = ∆Error(s) (5.10)
(1 + sTΦ )
where, KΦ and TΦ are gain constant and time constant of TCPS.
5.3.1 TCPS control strategy
Error signal to TCPS can be any signal such as the control area frequency deviation or area
control error to control the TCPS phase shifter angle. If the frequency deviation of control area-
1 is sensed as error signal, it can be used as the control signal to the TCPS unit to control the
TCPS phase shifter angle which results in controlling the tie line power flow [107, 108]. Thus,
KΦ
∆φ (s) = ∆F1 (s) (5.11)
(1 + sTΦ )
and
2π T̃12 KΦ
∆P̃tie,12 (s) = [∆F1 (s) − ∆F2 (s)] + T̃12 ∆F1 (s) (5.12)
s (1 + sTΦ )
The total incremental tie line power flow between control area-1 and other control areas, ∆P̃tie,1(s)
may be given as:
∆P̃tie,1 (s) = ∆P̃tie,12 (s) (5.13)
5.3.2 Simulation of the power system model considering TCPS
The TAIPS model with MSPG considering TCPS is shown in Fig. 5.7, where each control area
comprises reheat-thermal, hydro and gas generating units and the two equal areas are intercon-
nected. The simulation of this interconnected power system in a new power system environment
is based on the concepts of considering variety of generators with their corresponding gener-
ation contribution factors in each control area [6, 89, 117] and TCPS [108, 109]. The system
parameter values are given in Tabel 5.4 . It is assumed that Rki = Ri . The nominal loading
of each control area is taken 1640 MW with the power generation scheduling and generation
contribution factors as given in Table 5.5.
TH-1171_07610212
89
Table. 5.4: Simulation Parameters of TAIPS model with MSPG considering TCPS
Parameters Value
Prt1 = Prt2 2000 MW
PL1 = PL1 1640 MW
f 60 Hz
H1 = H2 5 MW-s/MVA
D1 =D2 0.0137 pu MW/Hz
T̃12 0.0433
TPS1 =TPS2 12.1951 s
TSG1 =TSG2 0.08 s
TT 1 =TT 2 0.3 s
TCD1 = TCD2 0.2 s
B1 =B2 0.4303
a12 -1
KR1 = KR2 0.3
TR1 = TR2 10 s
TW 1 = TW 2 1s
TRS1 = TRS2 5s
TRH1 =TRH2 28.75 s
TGH1 = TGH2 0.2 s
XG1 = XG2 0.6 s
YG1 =YG2 1s
cg1 =cg2 1
bg1 = cg1 0.05 s
TF1 = TF2 0.23 s
TCR1 = TCR2 0.01s
α11 = α12 0.579268
α21 = α22 0.274390
α31 = α32 0.146342
TCPS parameters
KΦ = 1.5 rad/Hz TΦ = 0.1 s
Φmax = 100 Φmin = −100
TH-1171_07610212
90
The state space form is obtained for the proposed power system model. The power system
has 26 state variables. State variables x1 , x12 , x25 and x26 are taken as output feedback states.
The state equations can be organized in the state-space form as described by the equations (2.52)
and 2.38.


The optimum gains of optimal OFC are obtained by running the MATLAB codes generated
on the basis of method described in section 2.6. The computer simulations are carried out
with the optimum controller gain settings. MATLAB control system toolbox [121] is used to
simulate the power system and to obtain dynamic responses of the system for 1% SLP in the
control area-1.
The optimum value of the K̃ for the OFC by minimizing the cost function for the power system
corresponding to nominal system parameters is
 
0.4202 −0.0936 0.3233 0.0808
−0.1634 −0.0039 −0.0386 −0.0522
 
 
 
0.1622 −0.0597 0.1340 0.0207
K̃ = 
 

 0.0719 0.8219 −0.1455 0.3840 
 
 −0.0469 −0.6530 −0.1181 −0.4162 
0.0473 0.3588 −0.0074 0.2008
and for the power system with TCPS is

 
1.4167 −0.3404 5.8889 −0.5138
−0.3702 0.0300 −1.1732 0.1551
 
 
 
0.4758 −0.2213 2.1480 −0.2559
K̃ = 
 

 0.0619 0.4614 0.5985 0.5440 
 
 −0.3225 −0.5033 −1.5146 −0.2861 
−0.0260 0.0202 0.1433 0.0902
TH-1171_07610212
91
Dynamic responses of the system are obtained for 1% SLP in the control area-1. The fre-
quency deviation responses of control area-1 and control area-2 are shown in Figs.5.9 and 5.10,
respectively. The tie line power deviation response is shown in Fig. 5.11. It is observed that
the OFC considering TCPS in power system gives better dynamic responses having relatively
smaller peak overshoot and lesser settling time with zero steady state error as compared to the
power system without TCPS. The quantitative comparison is made in Tables 5.6 and 5.7 where
percentage reduction in the peak OS of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 34.56, 50 and 39.13,
respectively and the settling time of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 41.61, 54.89 and 53.10,
respectively. The phase angle deviation of TCPS in response to 1% SLP in control area-1 is
shown in Fig. 5.12, where maximum phase angle deviation on positive side is 1.320 and on
negative side is 1.780 . Further, system dynamic responses are obtained for the wide range of
SLP varying from 1% to 4% in either control area and shown in Fig. 5.13. It is evident that
for SLP varying from 1% to 4%, the first peak overshoot increases with increase in the level of
SLP and settling time remains approximately same with zero steady state error. The dynamic
responses are improved with TCPS and satisfy the LFC requirements.
Table. 5.5: Power Generation Scheduling to match the nominal load of the individual control
area
Total control area Thermal Gas Hydro

load contribution contribution contribution
(MW) (MW) (MW) (MW)
1640 950 240 450
Generation contribution factor 0.579268 0.146342 0.274390
Table. 5.6: Dynamic response comparison in terms of peak OS
Peak OS of ∆ f1 (Hz) Peak OS of ∆ f1 (Hz) Peak OS of ∆P̃tie,1 (pu MW)

Without TCPS -0.0324 -0.0230 -0.0069
With TCPS -0.0212 -0.0115 -0.0042
% reduction in Peak OS 34.56 50 39.13
TH-1171_07610212
92
0.02
Without TCPS
with TCPS
0.01
−0.01
−0.02
−0.03
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.9: Frequency deviation response of control area-1
0.015
Without TCPS
with TCPS
0.01
0.005
−0.005
−0.01
−0.015
−0.02
−0.025
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.10: Frequency deviation response of control area-2
TH-1171_07610212
93
−3
x 10
3
Without TCPS

2 with TCPS
−1
−2
−3
−4
−5
−6
−7
0 5 10 15 20 25 30 35 40 45 50
time, s
Figure 5.11: Tie line power deviation response
0.04
TCPS phase angle deviation (rad)
0.03
0.02
0.01
−0.01
−0.02
−0.03
−0.04
0 5 10 15 20 25 30
time, s
Figure 5.12: TCPS phase shift deviation response
TH-1171_07610212
94
0.08
1% SLP
2% SLP
0.06
3% SLP
4% SLP
0.04
0.02
−0.02
−0.04
−0.06
−0.08
0 5 10 15 20 25 30
time, s
0.05
1% SLP
0.04 2% SLP
3% SLP
0.03 4% SLP
0.02
0.01
−0.01
−0.02
−0.03
−0.04
−0.05
0 5 10 15 20 25 30
time, s
1% SLP
2% SLP
3% SLP
0.01
4% SLP
−0.01
−0.02
0 5 10 15 20 25 30
time, s
Figure 5.13: Frequency deviation and tie line power responses for SLP varying from 1% to 4%
TH-1171_07610212
95
ST of ∆ f1 (s) ST of ∆ f1 (s) ST of ∆P̃tie,1 (s)

Without TCPS 32.27 40.26 49.37
With TCPS 18.84 18.16 23.15
% reduction in ST 41.61 54.89 53.10
5.4 Summary
An attempt is made in this chapter to study the the dynamic performance of LFC of the power
systems considering AC-DC tie lines and TCPS. A simple but practical controller is used to im-
prove the dynamic response of LFC system in the the new power system environment. The OFC
performs well for the power system with DC link giving better dynamic response having rela-
tively smaller peak overshoot and lesser settling time with zero steady state error as compared
to the power system considering AC tie lines only. The dynamic response of the system with
DC link is improved significantly with the percentage reduction in the peak OS of ∆ f1 , ∆ f2 and
∆Ptie,1 becoming 60.49, 90 and 44.92, respectively. Dynamic responses are obtained for wide
range of variation in SLP from 1% to 4% which satisfy the LFC requirements. The simulation
results show that proposed control strategy considering parallel AC-DC tie line is very effective
and guarantees good performance. Further, the OFC is proposed to control the TCPS phase
angle which in turn controls the tie line power flow. The proposed controller for the power
system with TCPS gives better dynamic responses having relatively smaller peak overshoot and
lesser settling time with zero steady state error as compared to the power system without TCPS.
The dynamic responses of the system with TCPS are improved significantly with the percentage
reduction in the peak OS of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 34.56, 50 and 39.13, respectively
and the settling time of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 41.61, 54.89 and 53.10, respectively.
Dynamic responses are obtained for wide range of variation in load disturbance from 1% to
4% which satisfy the LFC requirements. Simulation results show that the due to the presence
of TCPS, the dynamic performance in terms of settling time and peak overshoot is greatly im-
proved. The system with TCPS is capable of suppressing the control area frequency and tie line
power deviations more effectively under the occurrence of control area load perturbations. The
simulation results show that proposed control strategy considering TCPS is very effective and
TH-1171_07610212
96
guarantees good performance. Hence for all practical purposes, the controller is quite robust.
TH-1171_07610212
97
C HAPTER 6
LFC IN RESTRUCTURED POWER SYSTEM
ENVIRONMENT
6.1 Introduction
LFC is one of the important control problems in electric power system design and operation.
Any deviations in frequency can directly impact on power system operation and system re-
liability. A large frequency deviation can cause an unstable condition for the power system.
Maintaining frequency and tie line power interchanges with neighboring control areas at the
scheduled values are the two main primary objectives of a power system LFC [2–4, 90]. Many
researchers presented the LFC scheme of conventional power system using different control
strategies [1, 9, 13, 90–94, 108, 124].

In the restructured power system environment, vertically integrated system of conventional
power system do not exist [20]. In a competitive electricity market, generating companies
(GENCOs), distribution companies (DISCOs), transmission companies (TRANSCOs), and power
system operator (PSO) [5, 6, 11, 18, 19, 84] are all market players. As there are so many GEN-
COs and DISCOs in the restructured power system, a DISCO has the freedom to have a contract
with any GENCOs for the transaction of power. A DISCO of one control area can make con-
tract with a GENCO in another control area [19,20]. For stable and secure operation of a power
system, the PSO has to provide a number of ancillary services. One of the ancillary services
is the frequency regulation based on the concept of the LFC. The crucial role of LFC system
will continue in restructured power system environment with some modifications accounting
TH-1171_07610212
98
Chapter 6 LFC in restructured power system environment
bilateral transactions and deregulation policy [5].
An LFC system required for Poolco-based transactions described in [20, 112] utilizes an
integral controller. A method to find optimal controller gains of this type of controller for a two-
area system is proposed in [20]. In [115], B. Tyagi et. al proposed a general model for multi-area
LFC suitable for a competitive electricity environment. LFC work in restructured power system
is reported in [5, 11, 18–20, 84, 100] where they have considered either thermal or hydro system
in a control area. In new power system environment, a control area may have variety of sources
like hydro, thermal, gas, renewable etc., therefore representing a control area by thermal or
hydro system dynamics only may not result in a good design of LFC system [5,89,90]. Recently
some researchers studied the LFC of conventional power system considering hydro, thermal and
gas generating units in each control area [89, 90, 116, 117], however they did not consider the
LFC scheme in restructured power system environment.
Having studied the LFC scheme in conventional power systems with optimal OFC in the pre-
vious chapters, LFC scheme is presented in restructured power system environment. A TAIPS
with MSPG is proposed for LFC in restructured power system environment. Each control area
includes the more feasible multi-source combination of hydro, reheat thermal and gas generat-
ing units. An extensive analysis is done for LFC scheme considering Poolco-based transactions,
bilateral transactions, and a combination of these two transactions. The MATLAB simulation
results are authenticated by comparing with calculated(desired) values. A state space model of
the proposed power system in restructured power system environment is utilized for controller
design considering all the interface variables like control area frequencies, tie line power and all
other possible contracts between GENCOs and DISCOs. In this chapter, it is established that
the optimal OFC performs well in restructured power system environment also.
6.2 LFC in restructured power system environment
The LFC in a restructured power market should be designed to accommodate all possible trans-
actions [20, 112], such as Poolco-based transactions, bilateral transactions, and a combination
TH-1171_07610212
99
of these two transactions. In bilateral transactions, any DISCOs have the freedom to have a
contract with any GENCOs in its own and other control areas, whereas in Poolco-based transac-
tions, GENCOs participate in LFC of their own control areas only [19,20,100]. In a competitive
electricity market, Poolco and bilateral transactions may take place simultaneously. A DISCO
in any of the control areas and GENCOs in the same or in a different control area may also
negotiate bilateral contracts. The power market players are responsible for having a communi-
cation path to exchange contract data as well as measurements to perform the load following
function [20,84]. In such contracts, a GENCO changes its power output to supply the contracted
load as long as it does not exceed the contracted value. The DISCO is responsible to maintain
the load demand discipline as per contractual agreement.

In order to meet the Poolco-based and bilateral transactions, a DISCO participation matrix
(DPM) is used [20] for better visualization. In a DPM, number of rows and columns are equal
to the number of GENCOs and DISCOs, respectively. Each entry in a DPM is defined as
contract participation factor. The cp fkl is contract participation factor between kth GENCO
and l th DISCO. The cp fkl corresponds to the fraction of the total load power contracted by a
DISCO-l from a GENCO-k [20]. It is noted that ∑ cp fkl = 1. DPM shows the participation of
k
a DISCO in contract with a GENCO.
Each control area in this study is having three GENCOs and two DISCOs as shown in Fig.
6.1. Let GENCO-1, GENCO-2, GENCO-3, DISCO-1 and DISCO-2 be in control area-1 and
GENCO-4, GENCO-5, GENCO-6, DISCO-3 and DISCO-4 be in control area-2. As a particular
set of GENCOs is supposed to follow the total load demanded by a DISCO, information signals
must flow from the DISCO to the particular GENCO(s) specifying corresponding demands.
There may be uncontracted loads also in the control areas [84].
TH-1171_07610212
100
cpf11 + + cpf12
cpf21 cpf22
pu load + + + + cpf32 pu load
cpf31
DISCO-1 cpf42 DISCO-2
cpf41
cpf51 + + + + cpf52
cpf61
+ + cpf62
1 5 2 3 6 4
1/R1 Control area-1

7
1 GENCO-1
B1
+ -
+ 1 x4 (1+sKR1TR1) x3 1 x2 !m1(s) !D1(s)
u1 epf1 (1+sTSG1) (1+sTR1) (1+sTT1)
+ 2
+ x5 + - x1
1 x7 (1+sTRS1) x6 (1-sTW1) !m2(s) + + KPS1
+ (1+sTGH1)
u2 epf2 (1+sTRH1) (1+0.5sTW1) (1+sTPS1) !1(s)
+ + + -
x24 - GENCO-2
1 1/R1 8
s
+
+ 1 x11 (1+sXG1) x10 x9 1 x8 !m3(s)
u3 (1-sTCR1) Load demand of
epf3 (cg1+bg1s) (1+sYG1)
+ + (1+sTF1) (1+sTCD1) DISCOs (area-1) to
- GENCOs (area-2)
9 3 GENCO-3
1/R1
!tie12, error (s) + + !tie,12(s) +
x23 2. !"12
s
- -
Load demand of
DISCOs (area-2) to
4 1/R2 Control area-2 GENCOs (area-1)
a12 10 GENCO-4
+ - a12
x15 (1+sKR2TR2) x14 x13 !m4(s)
+ 1 1
u4 epf4 (1+sTSG2) (1+sTR2) (1+sTT2)
+ 5
+ + - x12
x18 x17 !m5(s) + KPS2
+ 1 (1+sTRS2) (1-sTW2) x16 +
u5 epf5 (1+sTGH2) (1+sTRH2) (1+0.5sTW2) (1+sTPS2) !2(s)
+
+ + -
x25 -
1 11 GENCO-5
1/R2
s !D2(s)
+
+ 1 x22 (1+sXG2) x21 x20 1 x19 !m6(s)
u6 epf6 (1-sTCR2)
(cg2+bg2s) (1+sYG2) (1+sTF2) (1+sTCD2)
B2 + +
6
-
12 GENCO-6
1/R2
11 7 8 9 12 10
cpf13 + + cpf14
cpf23 cpf24
pu load + + + + cpf34 pu load
cpf33
DISCO-3 cpf44 DISCO-4
cpf43
cpf53 + + + + cpf54
cpf63
+ + cpf64
Figure 6.1: A TAIPS with MSPG in restructured power system environment
TH-1171_07610212
101
The corresponding DPM will become

 
 cp f11 cp f12 cp f13 cp f14 
 
 cp f cp f22 cp f23 cp f24 
 21 
 
 cp f31 cp f32 cp f33 cp f34 
 
DPM =  

 (6.1)
 cp f41 cp f42 cp f43 cp f44 
 
 
 cp f cp f52 cp f53 cp f54 
 51 
 
cp f61 cp f62 cp f63 cp f64
The scheduled steady state power flow on the tie line [5] is given as
∆Ptie12,sch = (Pexp1 ) − (Pimp1) (6.2)
where;
Total power exported from control area-1 (Pexp1 ) is the demand of DISCOs in control area-2
from GENCOs in control area-1

and total power imported to control area-1 (Pimp1 ) is the demand of DISCOs in control area-1
from GENCOs in control area-2.
The error in tie line power (∆Ptie,12,error ) may be defined as
∆Ptie12,error = ∆Ptie,12 − ∆Ptie,12,sch (6.3)
∆Ptie,12,error vanishes in the steady state as the actual tie line power flow (∆Ptie,12 ) reaches the
scheduled tie line power flow.
In the restructured power system, the area control errors may be defined as:
ACE1 = B1 ∆ f1 + ∆Ptie,12,error (6.4)
ACE2 = B2 ∆ f2 + a12 ∆Ptie,12,error (6.5)
where, ACE1 and ACE2 are the area control errors of control area-1 and 2 respectively, B1 and
B2 are frequency bias parameters of control area-1 and 2 respectively, and a12 is control area
capacity ratio.
TH-1171_07610212
102
6.3 Simulation and results
A TAIPS with MSPG is simulated in restructured power system environment as shown in Fig.
6.1. GENCO-1 and GENCO-4 are thermal, GENCO-2 and GENCO-5 are hydro, and GENCO-
3 and GENCO-6 are gas generating units. The system parameters are given in Table 6.1. The
nominal load of each control area is taken 1640 MW for all the contracts. It is assumed that
Rki = Ri . The ep fk is the economic participation factor of kth GENCO. The economic partic-
ipation factor of a GENCO depends on its profile, participation, bid price and capacity in the
frequency regulation market [84]. The control signal (uk ) is the input to economic participation
factor block of kth GENCO.
The state space form is obtained for the proposed power system model. State variables x1 ,
x12 , x24 and x25 are taken as output feedback states. The state equations can be organized in the
state-space form as described by the equation (6.6) given below and equation (2.38).
ẋ = Ãx + B̃u + F̃w + F̄ w̄ (6.6)

control vector, u = [u1 u2 u3 u4 u5 u6 ]T
and the disturbance vector (w̄), due to contracts between GENCOs and DISCOs, is defined as:
w̄ = [∆P̃L1 ∆P̃L2 ∆P̃L3 ∆P̃L4 ]T

The constant matrices Ã, B̃ and C̃, the disturbance matrices F̃ and F̄, and design matrices Q̃
and R̃ are given in Appendix A.4.
It is assumed that all the GENCOs participate in the LFC with their corresponding economic
participation factor values ep f1 =0.6, ep f2 =0.3, ep f3 = 1 − (ep f1 + ep f2 ), ep f4 =0.6, ep f5 =0.3

and ep f6 = 1 − (ep f4 + ep f5 ). The total local load of ith control area(∆PDi ) is the sum of
contracted and uncontracted load demands of the DISCOs of ith control area only. The opti-
mum values of controller gains are obtained by running the MATLAB codes based on method
described in the section 2.6. The computer simulations are carried out with the optimum con-
TH-1171_07610212
103
troller gains. MATLAB control system toolbox [121] is used to simulate the power system
model and to obtain dynamic responses of the system for different contract scenarios (case-1, 2
and 3).
Table. 6.1: Simulation parameters of TAIPS with MSPG in restructured power environment
Parameters Value
Prt1 = Prt2 2000 MW
PL1 = PL1 1640 MW
f 60 Hz
H1 = H2 5 MW-s/MVA
D1 =D2 0.0137 pu MW/Hz
T12 0.0433
TPS1 =TPS2 12.1951 s
TSG1 =TSG2 0.08 s
TT 1 =TT 2 0.3 s
TCD1 = TCD2 0.2 s
B1 =B2 0.4303
a12 -1
KR1 = KR2 0.3
TR1 = TR2 10 s
TW 1 = TW 2 1s
TRS1 = TRS2 5s
TRH1 =TRH2 28.75 s
TGH1 = TGH2 0.2 s
XG1 = XG2 0.6 s
YG1 =YG2 1s
cg1 =cg2 1
bg1 = cg2 0.05 s
TF1 = TF2 0.23 s
TCR1 = TCR2 0.01s
Case-1: Poolco based transactions
In this kind of transactions, GENCOs participate in LFC of their own control areas only [19,20].
It is assumed that the load change occurs only in control area-1. Thus, the load is demanded
only by DISCO-1 and DISCO-2. Let the value of this load demand be 0.05 pu MW for each of
them. DISCO-1 and DISCO-2 demand equally from GENCO-1, GENCO-2 and GENCO-3. A
case of Poolco based contracts between DISCOs and available GENCOs is simulated based on
TH-1171_07610212
104
the following DPM-

 
 0.3333 0.3333 0 0 
 
 0.3333 0.3333 0 0 
 
 
 0.3333 0.3333 0 0 
 
DPM = 




 0 0 0 0 

 

 0 0 0 0 

 
0 0 0 0
DISCO-3 and DISCO-4 do not demand power from any GENCOs, and hence the corresponding
contract participation factors are zero. Fig. 6.2 shows control area frequency deviations and
actual tie line power deviation, following a step change in the load demands of DISCO-1 and
DISCO-2. The scheduled steady state tie line power flow is zero. The actual power on the tie
line settles to zero. The frequency and tie line power deviation in each control area settles to
zero in the steady state. In the steady state, generation of a GENCO must match the demand
of DISCOs in contract with it. The desired generation of the kth GENCO in pu MW can be
expressed [20] in terms of contract participation factors and the total contracted demand of
DISCOs as-
∆Pmk = cp fk1 ∆P̃L1 + cp fk2 ∆P̃L2 + cp fk3 ∆P̃L3 + cp fk4 ∆P̃L4 (6.7)
Where ∆P̃L1 , ∆P̃L2, ∆P̃L3 and ∆P̃L4 are the total contracted demands of DISCO-1, 2, 3 and 4
respectively.
In this case ∆P̃L1 =0.05 pu MW, ∆P̃L2 =0.05 pu MW, ∆P̃L3=0 and ∆P̃L4 =0 and contract par-
ticipation factors are as given in DPM. From the equation (6.7), the calculated values (desired
values) of generations of GENCOs are ∆Pm1 =0.0333 pu MW, ∆Pm2 =0.0333 pu MW, ∆Pm3 =
0.0333 pu MW, ∆Pm4 =0, ∆Pm5 =0 and ∆Pm6 =0 at steady state. The generated powers (actual
powers) of various GENCOs in response to contract with DISCOs are given in Fig. 6.3. The
actual generated powers of the GENCOs reach the desired values in the steady state. GENCOs
of control area-2 do not transact the power; hence, their change in generated power is zero at
steady state.
TH-1171_07610212
105
case−1
0.15
0.1
0.05
−0.05
−0.1
−0.15
−0.2
0 20 40 60 80 100 120 140
time, s

case−1
0.06
0.04
0.02
−0.02
−0.04
−0.06
−0.08
0 20 40 60 80 100 120 140
time, s

case−1
0.01
Deviation in actual tie line power(pu MW)
0.005
−0.005
−0.01
−0.015
−0.02
−0.025
−0.03
0 20 40 60 80 100 120 140
time, s
Figure 6.2: Frequency and tie line power deviation responses for case-1
TH-1171_07610212
106
case−1
Deviation in power outputs of GENCOs,area−1 (pu MW)

0.1
GENCO−1
GENCO−2
GENCO−3
0.08
0.06
0.04
0.02
−0.02
0 20 40 60 80 100 120 140
time, s
(a) Generator power output response of control area-1
case−1
0.025
GENCO−4
GENCO−5
0.02 GENCO−6
0.015
0.01
0.005
−0.005
−0.01
−0.015
0 20 40 60 80 100 120 140
time, s
(b) Generator power output response of control area-2
Figure 6.3: Generator power output response for case-1
TH-1171_07610212
107
Case-2: Poolco and bilateral transactions
In this kind of transactions, a DISCO has the freedom to have a contract with any GENCOs in
its own and other control areas [20] [19]. A case of combinations of Poolco and bilateral based
contracts between DISCOs and available GENCOs is simulated based on the following DPM-
 
 0.2 0.1 0.3 0 
 
 0.2 0.2 0.1 0.1666 
 
 
 0.1 0.3 0.1 0.1666 
 
DPM =  


 0.2 0.1 0.1 0.3336 
 
 
 0.2 0.2 0.2 0.1666 
 
 
0.1 0.1 0.2 0.1666
It is assumed that each DISCO demands 0.05 pu MW power from GENCOs as defined
by contract participation factors in DPM matrix. The power system in Fig. 6.1 is simulated
using this data and the results are shown in Figs.6.4 and 6.5. In this case ∆P̃L1=0.05 pu MW,
∆P̃L2 =0.05 pu MW, ∆P̃L3 =0.05 pu MW and ∆P̃L4=0.05 pu MW and contract participation factors
are as given in DPM. The calculated values(desired values) of generations of GENCOs are
∆Pm1 = 0.0300 pu MW, ∆Pm2 =0.0333 pu MW, ∆Pm3=0.0333 pu MW, ∆Pm4 =0.0367 pu MW,
∆Pm5 =0.0383 pu MW and ∆Pm6 = 0.0283 pu MW at steady state. The generated powers (actual
powers) of various GENCOs in response to contract with DISCOs are shown in Fig. 6.5. As
shown in Fig. 6.4(a-b), the control area frequency deviations settle to zero value. The tie line
power deviation settles to scheduled value -0.0033 pu MW at steady state as shown in Fig.
6.4(c) which matches with the calculated (desired) value -0.0033 pu MW.
Case-3: contract violation
In some situations, DISCOs may violate a contract by demanding excess power. This excess
power (uncontracted power) must be supplied by the GENCOs in the same control area as the
DISCO [19, 20]. This kind of demand is reflected as a local load of the control area. Consider
case-1 again with a modification that DISCO-1 demands 0.01 pu MW of excess power. The
TH-1171_07610212
108
case−2
0.05
0
−0.05
−0.1
−0.15
−0.2
0 20 40 60 80 100 120 140
time, s
case−2
0.05
−0.05
−0.1
−0.15
−0.2
0 20 40 60 80 100 120 140
time, s
case−2
0.015
0.01
0.005
−0.005
−0.01
0 20 40 60 80 100 120 140
time, s
TH-1171_07610212
109
case−2
0.12
GENCO−1
GENCO−2
0.1
GENCO−3
0.08
0.06
0.04
0.02
−0.02
−0.04
0 20 40 60 80 100 120 140 160 180 200
time, s

case−2
0.1
GENCO−4
GENCO−5
0.08 GENCO−6
0.06
0.04
0.02
−0.02
−0.04
0 20 40 60 80 100 120 140 160 180 200
time, s
TH-1171_07610212
110
total local load in control area-1 becomes
∆PD1 =(Load of DISCO-1)+(Load of DISCO-2) =(0.05+0.01)+(0.05)=0.11 pu MW

Similarly, the total local load in control area-2 becomes
∆PD2 =(Load of DISCO-3)+(Load of DISCO-4) =(0)+(0)=0 pu MW
The frequency and tie line power deviations vanish in the steady state, shown in Fig. 6.6.
As DPM is the same as in case-1 and the excess load is taken up by GENCOs in the same
control area, the tie line power is the same as in case-1 in steady state. As shown in Fig.6.7,
the generation of GENCOs-4, 5 and 6 is not affected by the excess load of DISCO-1. The
uncontracted load of DISCO-1 is reflected in the generations of GENCO-1, 2 and 3. Sum of
power generations of GENCOs in control area-1 as shown in Fig.6.7 matches with calculated
(desired) value, 0.11 pu MW.
TH-1171_07610212
111
case−3
0.15
0.1
0.05
−0.05
−0.1
−0.15
−0.2
0 20 40 60 80 100 120 140
time, s
case−3
0.06
0.04
0.02
−0.02
−0.04
−0.06
−0.08
0 20 40 60 80 100 120 140
time, s

case−3
0.01
0.005
−0.005
−0.01
−0.015
−0.02
−0.025
−0.03
0 20 40 60 80 100 120 140
time, s
TH-1171_07610212
112
case−3
0.12
Deviation in power outputs of GENCOs,area−1 (pu MW) GENCO−1
GENCO−2
0.1 GENCO−3
0.08
0.06
0.04
0.02
−0.02
0 20 40 60 80 100 120 140 160 180 200
time, s

case−3
0.03
GENCO−4
0.025 GENCO−5
GENCO−6
0.02
0.015
0.01
0.005
−0.005
−0.01
−0.015
0 20 40 60 80 100 120 140 160 180 200
time, s
TH-1171_07610212
113
6.4 Summary
A new model of TAIPS with MSPG in restructured power system environment is proposed.
Modified LFC scheme is presented in restructured power system environment. A state space
model of the proposed power system in restructured power system environment is utilized for
controller design considering all the interface variables like control area frequencies, tie line
power and all other possible contracts between GENCOs and DISCOs. An extensive analysis
is done for LFC scheme considering Poolco-based transactions, bilateral transactions, and a
combination of these two transactions. It is found that in all the cases, the area frequency
error becomes zero in the steady state which satisfy the LFC requirements. It is found that
actual values of generations and tie-line power exchanges of GENCOs obtained by MATLAB
Simulink model of the power system are matching with the corresponding calculated (desired)
values.
TH-1171_07610212
114
C HAPTER 7
C ONCLUSIONS AND F UTURE WORK
7.1 Concluding remarks
Many research work on LFC have been reported in recent years. Still there is much room for
further improvement and extensions of LFC strategies. This thesis presents some improved
results on LFC study of conventional and restructured power systems emphasizing on MSPG.
New LFC models of the power systems with MSPG have been proposed. OFC is presented with
pragmatic point of view. Briefly the results are summarized as follows:
A. Mathematical modeling and LFC of power systems
Mathematical modeling of the SAPS, MAIPS and MAIPS with MSPG are presented for
LFC study. Keeping in view the new power system environment, the concept of MSPG in a
control area is introduced. In addition to LFC models of SAPS and MAIPS, a new model of
MAIPS with MSPG is presented for LFC study. The OFC is introduced. The control system
design equations and algorithms have been presented.
B. LFC of conventional power systems

The LFC of various conventional TAIPSs has been studied using OFC. The combinations of
non-reheat, reheat and hydro turbines are taken to demonstrate the suitability of OFC for LFC
of interconnected power systems. The dynamic responses are obtained for TAIPSs considering
(i) thermal-thermal with non-reheat turbine, (ii) thermal-thermal with reheat turbine and (iii)
hydro-thermal generating units. The Dynamic responses obtained with OFC have been com-
pared with that of FSFC. It has been found that dynamic responses obtained with proposed OFC
are comparatively better and competes well with FSFC, satisfying the LFC requirements. The
TH-1171_07610212
115
Chapter 7 Conclusions and future work
OFC is presented from pragmatic point of view which uses less number of states as feedback,
thereby reducing the controller cost and complexities.

C. LFC of power systems with multi-source power generation
LFC scheme is presented for power systems with MSPG. The performance of the proposed
controller is demonstrated on the multi-source power systems and its dynamic responses are
compared with that of FSFC. The OFC gives better dynamic response having relatively smaller
peak overshoot and lesser settling time with zero steady state error as compared to that of FSFC.
The effect of GRC on dynamic response is discussed. The critical examination of dynamic re-
sponses reveals that GRC results in increased peak overshoot, however settling time remains
same for this particular power system model. The controller gives satisfactory dynamic re-
sponses for turbines considering GRC. It shows that OFC is able to accommodate the effect
of GRC also. Investigations reveal that it is better to prefer the medium value of R i.e. 4%
with corresponding optimum controller gains to provide better dynamic response of LFC for
the proposed system. The dynamic responses are obtained for 1% to 5% SLPs. It is observed
that for SLPs varying from 1% to 5%, the first peak overshoot increases with increase in the
level of SLP and settling time remains approximately same with zero steady state error. The
sensitivity analysis reveals that the optimum values of controller gains obtained for nominal
system parameters and load condition are quite insensitive to wide parameter variation ±25%.
Hence for all practical purposes, the controller is quite robust. The LFC of hydro power plants
operational in Iran has also been studied. The proposed controller performs well on this sys-
tem and improves the frequency deviation responses remarkably. The simulation results show
that proposed control strategy is very effective and guarantees good performance. In fact, this
method provides a control system that satisfies the load frequency control requirements with a
credible dynamic response.
D. LFC of power systems considering AC-DC tie lines and TCPS
An attempt is made to study the the dynamic performance of LFC of the power systems
considering AC-DC tie lines and TCPS. A simple but practical controller is used to improve the
dynamic response of LFC system in the the new power system environment. The OFC performs
TH-1171_07610212
116
well for the power system with DC link, giving better dynamic responses having relatively
smaller peak overshoot and lesser settling time with zero steady state error as compared to the
power system considering AC tie lines only. The dynamic response of the system with DC link
is improved significantly with the percentage reduction in the peak overshoot of ∆ f1 , ∆ f2 and
∆Ptie,1 becoming 60.49, 90 and 44.92, respectively. Dynamic responses are obtained for wide
range of variation in SLP from 1% to 4% which satisfy the LFC requirements. The simulation
results show that proposed control strategy considering parallel AC-DC tie line is very effective
and guarantees good performance. Further, the OFC is proposed to control the TCPS phase
angle which in turn controls the tie line power flow. The proposed controller for the power
system with TCPS gives better dynamic responses having relatively smaller peak overshoot and
lesser settling time with zero steady state error as compared to the power system without TCPS.
The dynamic response of the system with TCPS is improved significantly with the percentage
reduction in the peak OS of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 34.56, 50 and 39.13, respectively
and the settling time of ∆ f1 , ∆ f2 and ∆P̃tie,1 becoming 41.61, 54.89 and 53.10, respectively.
Dynamic responses are obtained for wide range of variation in load disturbance from 1% to
4% which satisfy the LFC requirements. Simulation results show that the due to the presence
of TCPS, the dynamic performance in terms of settling time and peak overshoot is greatly
improved. The system with TCPS is capable of suppressing the control area frequency and tie
line power deviations more effectively under the occurrence of control area load perturbations.
The simulation results show that proposed control strategy considering TCPS is very effective
and guarantees good performance. Hence for all practical purposes, the controller is quite
robust.
E. LFC in restructured power system environment

A new model of TAIPS with MSPG in restructured power system environment is proposed.
Modified LFC scheme is presented in restructured power system environment. A state space
model of the proposed power system in restructured power system environment is utilized for
controller design considering all the interface variables like control area frequencies, tie line
power and all other possible contracts between GENCOs and DISCOs. An extensive analysis
TH-1171_07610212
117
is done for LFC scheme considering Poolco-based transactions, bilateral transactions, and a
combination of these two transactions. It is found that in all the cases, the area frequency
error becomes zero in the steady state which satisfy the LFC requirements. It is found that
actual values of generations and tie-line power exchanges of GENCOs obtained by MATLAB
Simulink model of the power system are matching with the corresponding calculated (desired)
values. Thus, the proposed LFC scheme is suitable in restructured power system environment.
7.2 Suggestions for further work
Following the LFC described in this thesis, a number of possible directions for extensions to
this work are discussed below:
• Research work may be extended to renewable energy sources integration in the control
area.
• Method may be extended to LFC study of the Micro-grids.
• Some energy storage devices may be included.
• Some evolutionary algorithms may be incorporated with proposed method and power
system models.
• LFC study may be extended for combined cycle gas turbine and pumped storage hydro
power plants.
• Price based LFC mechanism can be extended to improve frequency profile of the power
system.
TH-1171_07610212
118
A PPENDIX A
M ATRICES OF POWER SYSTEMS
A.1 Matrices of TAIPS with MSPG

The matrix Ã is of the dimension 25 × 25, where Ãi, j is the element of ith row and jth column of the matrix Ã. The non-zero elements are given
below:
Ã1,1 = −1/TPS1
Ã1,2 = KPS1 α11 /TPS1
Ã1,5 = KPS1 α21 /TPS1
Ã1,8 = KPS1 α31 /TPS1
Ã1,23 = −KPS1 /TPS1
Ã2,2 = −1/TT 1
Ã2,3 = 1/TT 1
Ã3,1 = −KR1 /(TSG1 R11 )
Ã3,3 = −1/TR1
Ã3,4 = 1/TR1 − KR1 /TSG1
Ã4,1 = −1/(TSG1 R11 )
Ã4,4 = −1/TSG1
Ã5,1 = 2TRS1 /(TRH1 TGH1 R21 )
Ã5,5 = −2/TW 1
Ã5,6 = 2/TW 1 + 2/TRH1
Ã5,7 = 2TRS1 /(TRH1 TGH1 ) − 2/TRH1
Ã6,1 = −TRS1 /(TRH1 TGH1 R21 )
Ã6,6 = −1/TRH1
Ã6,7 = 1/TRH1 − TRS1 /(TRH1 TGH1 )
Ã7,1 = −1/(TGH1 R21 )
Ã7,7 = −1/TGH1
Ã8,8 = −1/TCD1
Ã8,9 = 1/TCD1
Ã9,1 = TCR1 XG1 /(TF1 R31YG1 bg1 )
Ã9,9 = −1/TF1
Ã9,10 = 1/TF1 + TCR1 /(TF1YG1 )
Ã9,11 = TCR1 XG1 cg1 /(TF1YG1 bg1 ) − TCR1 /(TF1YG1 )
TH-1171_07610212
119
Appendix A
Ã10,1 = −XG1 /(R31YG1 bg1 )

Ã10,10 = −1/YG1
Ã10,11 = 1/YG1 − XG1 cg1 /(YG1 bg1 )
Ã11,1 = −1/(R31 bg1 )
Ã11,11 = −cg1 /bg1
Ã12,12 = −1/TPS2
Ã12,13 = KPS2 α12 /TPS2
Ã12,16 = KPS2 α22 /TPS2
Ã12,19 = KPS2 α32 /TPS2
Ã12,23 = −a12 KPS2 /TPS2
Ã13,13 = −1/TT 2
Ã13,14 = 1/TT 2
Ã14,12 = −KR2 /(TSG2 R12 )
Ã14,14 = −1/TR2
Ã14,15 = 1/TR2 − KR2 /TSG2
Ã15,12 = −1/(TSG2 R12 )
Ã15,15 = −1/TSG2
Ã16,12 = 2TRS2 /(TRH2 TGH2 R22 )
Ã16,16 = −2/TW 2
Ã16,17 = 2/TW 2 + 2/TRH2
Ã16,18 = 2TRS2 /(TRH2 TGH2 ) − 2/TRH2
Ã17,12 = −TRS2 /(TRH2 TGH2 R22 )
Ã17,17 = −1/TRH2
Ã17,18 = 1/TRH2 − TRS2 /(TRH2 TGH2 )
Ã18,12 = −1/(TGH2 R22 )
Ã18,18 = −1/TGH2
Ã19,19 = −1/TCD2
Ã19,20 = 1/TCD2
Ã20,12 = TCR2 XG2 /(TF2 R32YG2 bg2 )
Ã20,20 = −1/TF2
Ã20,21 = 1/TF2 + TCR2 /(TF2YG2 )
Ã21,12 = −XG2 /(R32YG2 bg2 )
Ã21,21 = −1/YG2
Ã21,22 = 1/YG2 − XG2 cg2 /(YG2 bg2 )
Ã22,12 = −1/(R32 bg2 )
Ã22,22 = −cg2 /bg2
Ã23,1 = 2π T12
Ã23,12 = −2π T12
Ã24,1 = B1
Ã24,23 = 1
Ã25,12 = B2
Ã25,23 = a12
TH-1171_07610212
120
Appendix A
The matrix B̃ is of the dimension 25 × 6, where B̃i, j is the element of ith row and jth column of the matrix B̃. The non-zero elements are given
below:
B̃3,1 = KR1 /TSG1
B̃4,1 = 1/TSG1
B̃5,2 = −2TRS1 /(TRH1 TGH1 )
B̃6,2 = TRS1 /(TRH1 TGH1 )
B̃7,2 = 1/TGH1
B̃9,3 = −XG1 TCR1 /(TF1YG1 bg1 )
B̃10,3 = XG1 /(YG1 bg1 )
B̃11,3 = 1/bg1
B̃14,4 = KR2 /TSG2
B̃15,4 = 1/TSG2
B̃16,5 = −2TRS2 /(TRH2 TGH2 )
B̃17,5 = TRS2 /(TRH2 TGH2 )
B̃18,5 = 1/TGH2
B̃20,6 = −XG2 TCR2 /(TF2YG2 bg2 )
B̃21,6 = XG2 /(YG2 bg2 )
B̃22,6 = 1/bg2
The matrix F̃ is of the dimension 25 × 2, where F̃i, j is the element of ith row and jth column of the matrix F̃. The non-zero elements are given
below:
F̃1,1 = −KPS1 /TPS1
F̃12,2 = −KPS2 /TPS2
 
 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 
C̃ = 
 

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 
 
 
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
The matrix Q̃ is of the dimension 25 × 25, where Q̃i, j is the element of ith row and jth column of the matrix Q̃. The non-zero elements are given
below:
Q̃1,1 = 1 + B21
Q̃1,23 = B1
Q̃12,12 = 1 + B22
Q̃12,23 = B2 a12
Q̃23,1 = B1
Q̃23,12 = B2 a12
Q̃23,23 = 1 + a212
Q̃24,24 = 1
Q̃25,25 = 1
and the matrix R̃ is an identity matrix of the dimension 6 × 6.
TH-1171_07610212
121
Appendix A
A.2 Matrices of TAIPS with MSPG considering AC-DC tie
lines
below:
Ã1,1 = −1/TPS1
Ã1,2 = KPS1 α11 /TPS1
Ã1,5 = KPS1 α21 /TPS1
Ã1,8 = KPS1 α31 /TPS1
Ã1,23 = −KPS1 /TPS1
Ã1,24 = −KPS1 /TPS1
Ã2,2 = −1/TT 1
Ã2,3 = 1/TT 1
Ã3,1 = −KR1 /(TSG1 R11 )
Ã3,3 = −1/TR1
Ã3,4 = 1/TR1 − KR1 /TSG1
Ã4,1 = −1/(TSG1 R11 )
Ã4,4 = −1/TSG1
Ã5,1 = 2TRS1 /(TRH1 TGH1 R21 )
Ã5,5 = −2/TW 1
Ã5,6 = 2/TW 1 + 2/TRH1
Ã5,7 = 2TRS1 /(TRH1 TGH1 ) − 2/TRH1
Ã6,1 = −TRS1 /(TRH1 TGH1 R21 )
Ã6,6 = −1/TRH1
Ã6,7 = 1/TRH1 − TRS1 /(TRH1 TGH1 )
Ã7,1 = −1/(TGH1 R21 )
Ã7,7 = −1/TGH1
Ã8,8 = −1/TCD1
Ã8,9 = 1/TCD1
Ã9,1 = TCR1 XG1 /(TF1 R31YG1 bg1 )
Ã9,9 = −1/TF1
Ã9,10 = 1/TF1 + TCR1 /(TF1YG1 )
Ã10,1 = −XG1 /(R31YG1 bg1 )
Ã10,10 = −1/YG1
Ã10,11 = 1/YG1 − XG1 cg1 /(YG1 bg1 )
Ã11,1 = −1/(R31 bg1 )
Ã11,11 = −cg1 /bg1
Ã12,12 = −1/TPS2
Ã12,13 = KPS2 α12 /TPS2
Ã12,16 = KPS2 α22 /TPS2
Ã12,19 = KPS2 α32 /TPS2
TH-1171_07610212
122
Appendix A
Ã12,23 = −a12 KPS2 /TPS2

Ã12,25 = −KPS2 /TPS2
Ã13,13 = −1/TT 2
Ã13,14 = 1/TT 2
Ã14,12 = −KR2 /(TSG2 R12 )
Ã14,14 = −1/TR2
Ã14,15 = 1/TR2 − KR2 /TSG2
Ã15,12 = −1/(TSG2 R12 )
Ã15,15 = −1/TSG2
Ã16,12 = 2TRS2 /(TRH2 TGH2 R22 )
Ã16,16 = −2/TW 2
Ã16,17 = 2/TW 2 + 2/TRH2
Ã16,18 = 2TRS2 /(TRH2 TGH2 ) − 2/TRH2
Ã17,12 = −TRS2 /(TRH2 TGH2 R22 )
Ã17,17 = −1/TRH2
Ã17,18 = 1/TRH2 − TRS2 /(TRH2 TGH2 )
Ã18,12 = −1/(TGH2 R22 )
Ã18,18 = −1/TGH2
Ã19,19 = −1/TCD2
Ã19,20 = 1/TCD2
Ã20,12 = TCR2 XG2 /(TF2 R32YG2 bg2 )
Ã20,20 = −1/TF2
Ã20,21 = 1/TF2 + TCR2 /(TF2YG2 )
Ã21,12 = −XG2 /(R32YG2 bg2 )
Ã21,21 = −1/YG2
Ã21,22 = 1/YG2 − XG2 cg2 /(YG2 bg2 )
Ã22,12 = −1/(R32 bg2 )
Ã22,22 = −cg2 /bg2
Ã23,1 = 2π T12
Ã23,12 = −2π T12
Ã24,1 = KDC1 /TDC1
Ã24,24 = −1/TDC1
Ã25,12 = KDC2 /TDC2
Ã25,25 = −1/TDC2
Ã26,1 = B1
Ã26,23 = 1
Ã27,12 = B2
Ã27,23 = a12
below:
B̃3,1 = KR1 /TSG1
B̃4,1 = 1/TSG1
TH-1171_07610212
123
Appendix A
B̃5,2 = −2TRS1 /(TRH1 TGH1 )

B̃6,2 = TRS1 /(TRH1 TGH1 )
B̃7,2 = 1/TGH1
B̃9,3 = −XG1 TCR1 /(TF1YG1 bg1 )
B̃10,3 = XG1 /(YG1 bg1 )
B̃11,3 = 1/bg1
B̃14,4 = KR2 /TSG2
B̃15,4 = 1/TSG2
B̃16,5 = −2TRS2 /(TRH2 TGH2 )
B̃17,5 = TRS2 /(TRH2 TGH2 )
B̃18,5 = 1/TGH2
B̃20,6 = −XG2 TCR2 /(TF2YG2 bg2 )
B̃21,6 = XG2 /(YG2 bg2 )
B̃22,6 = 1/bg2
below:
F̃1,1 = −KPS1 /TPS1
F̃12,2 = −KPS2 /TPS2
 
 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
C̃ = 
 

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 
 
 
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
below:
Q̃1,1 = 1 + B21
Q̃1,23 = B1
Q̃12,12 = 1 + B22
Q̃12,23 = B2 a12
Q̃23,1 = B1
Q̃23,12 = B2 a12
Q̃23,23 = 1 + a212
Q̃26,26 = 1
Q̃27,27 = 1
TH-1171_07610212
124
Appendix A
A.3 Matrices of TAIPS with MSPG considering TCPS

below:
Ã1,1 = −1/TPS1
Ã1,2 = KPS1 α11 /TPS1
Ã1,5 = KPS1 α21 /TPS1
Ã1,8 = KPS1 α31 /TPS1
Ã1,23 = −KPS1 /TPS1
Ã1,24 = −KPS1 /TPS1
Ã2,2 = −1/TT 1
Ã2,3 = 1/TT 1
Ã3,1 = −KR1 /(TSG1 R11 )
Ã3,3 = −1/TR1
Ã3,4 = 1/TR1 − KR1 /TSG1
Ã4,1 = −1/(TSG1 R11 )
Ã4,4 = −1/TSG1
Ã5,1 = 2TRS1 /(TRH1 TGH1 R21 )
Ã5,5 = −2/TW 1
Ã5,6 = 2/TW 1 + 2/TRH1
Ã5,7 = 2TRS1 /(TRH1 TGH1 ) − 2/TRH1
Ã6,1 = −TRS1 /(TRH1 TGH1 R21 )
Ã6,6 = −1/TRH1
Ã6,7 = 1/TRH1 − TRS1 /(TRH1 TGH1 )
Ã7,1 = −1/(TGH1 R21 )
Ã7,7 = −1/TGH1
Ã8,8 = −1/TCD1
Ã8,9 = 1/TCD1
Ã9,1 = TCR1 XG1 /(TF1 R31YG1 bg1 )
Ã9,9 = −1/TF1
Ã9,10 = 1/TF1 + TCR1 /(TF1YG1 )
Ã10,1 = −XG1 /(R31YG1 bg1 )
Ã10,10 = −1/YG1
Ã10,11 = 1/YG1 − XG1 cg1 /(YG1 bg1 )
Ã11,1 = −1/(R31 bg1 )
Ã11,11 = −cg1 /bg1
Ã12,12 = −1/TPS2
Ã12,13 = KPS2 α12 /TPS2
Ã12,16 = KPS2 α22 /TPS2
Ã12,19 = KPS2 α32 /TPS2
Ã12,23 = −a12 KPS2 /TPS2
Ã12,24 = −a12 KPS2 /TPS2
Ã13,13 = −1/TT 2
TH-1171_07610212
125
Appendix A
Ã13,14 = 1/TT 2
Ã14,12 = −KR2 /(TSG2 R12 )
Ã14,14 = −1/TR2
Ã14,15 = 1/TR2 − KR2 /TSG2
Ã15,12 = −1/(TSG2 R12 )
Ã15,15 = −1/TSG2
Ã16,12 = 2TRS2 /(TRH2 TGH2 R22 )
Ã16,16 = −2/TW 2
Ã16,17 = 2/TW 2 + 2/TRH2
Ã16,18 = 2TRS2 /(TRH2 TGH2 ) − 2/TRH2
Ã17,12 = −TRS2 /(TRH2 TGH2 R22 )
Ã17,17 = −1/TRH2
Ã17,18 = 1/TRH2 − TRS2 /(TRH2 TGH2 )
Ã18,12 = −1/(TGH2 R22 )
Ã18,18 = −1/TGH2
Ã19,19 = −1/TCD2
Ã19,20 = 1/TCD2
Ã20,12 = TCR2 XG2 /(TF2 R32YG2 bg2 )
Ã20,20 = −1/TF2
Ã20,21 = 1/TF2 + TCR2 /(TF2YG2 )
Ã21,12 = −XG2 /(R32YG2 bg2 )
Ã21,21 = −1/YG2
Ã21,22 = 1/YG2 − XG2 cg2 /(YG2 bg2 )
Ã22,12 = −1/(R32 bg2 )
Ã22,22 = −cg2 /bg2
Ã23,1 = 2π T̃12
Ã23,12 = −2π T̃12
Ã24,1 = T̃12 Kφ /Tφ
Ã24,24 = −1/Tφ
Ã25,1 = B1
Ã25,23 = 1
Ã25,24 = 1
Ã26,12 = B2
Ã26,23 = a12
Ã26,24 = a12
below:
B̃3,1 = KR1 /TSG1
B̃4,1 = 1/TSG1
B̃5,2 = −2TRS1 /(TRH1 TGH1 )
B̃6,2 = TRS1 /(TRH1 TGH1 )
B̃7,2 = 1/TGH1
TH-1171_07610212
126
Appendix A
B̃9,3 = −XG1 TCR1 /(TF1YG1 bg1 )

B̃10,3 = XG1 /(YG1 bg1 )
B̃11,3 = 1/bg1
B̃14,4 = KR2 /TSG2
B̃15,4 = 1/TSG2
B̃16,5 = −2TRS2 /(TRH2 TGH2 )
B̃17,5 = TRS2 /(TRH2 TGH2 )
B̃18,5 = 1/TGH2
B̃20,6 = −XG2 TCR2 /(TF2YG2 bg2 )
B̃21,6 = XG2 /(YG2 bg2 )
B̃22,6 = 1/bg2
below:
F̃1,1 = −KPS1 /TPS1
F̃12,2 = −KPS2 /TPS2
 
 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
C̃ = 
 

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 
 
 
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
below:
Q̃1,1 = 1 + B21
Q̃1,23 = B1
Q̃1,24 = B1
Q̃12,12 = 1 + B22
Q̃12,23 = B2 a12
Q̃12,24 = B2 a12
Q̃23,1 = B1
Q̃23,12 = B2 a12
Q̃23,23 = 1 + a212
Q̃23,24 = 1 + a12
Q̃24,1 = B1
Q̃24,12 = B2 a12
Q̃24,23 = 1 + a12
Q̃25,25 = 1
Q̃26,26 = 1
TH-1171_07610212
127
Appendix A
A.4 Matrices of TAIPS with MSPG in restructured power
system
below:
Ã1,1 = −1/TPS1
Ã1,2 = KPS1 /TPS1
Ã1,5 = KPS1 /TPS1
Ã1,8 = KPS1 /TPS1
Ã1,23 = −KPS1 /TPS1
Ã2,2 = −1/TT 1
Ã2,3 = 1/TT 1
Ã3,1 = −KR1 /(TSG1 R11 )
Ã3,3 = −1/TR1
Ã3,4 = 1/TR1 − KR1 /TSG1
Ã4,1 = −1/(TSG1 R11 )
Ã4,4 = −1/TSG1
Ã5,1 = 2TRS1 /(TRH1 TGH1 R21 )
Ã5,5 = −2/TW 1
Ã5,6 = 2/TW 1 + 2/TRH1
Ã5,7 = 2TRS1 /(TRH1 TGH1 ) − 2/TRH1
Ã6,1 = −TRS1 /(TRH1 TGH1 R21 )
Ã6,6 = −1/TRH1
Ã6,7 = 1/TRH1 − TRS1 /(TRH1 TGH1 )
Ã7,1 = −1/(TGH1 R21 )
Ã7,7 = −1/TGH1
Ã8,8 = −1/TCD1
Ã8,9 = 1/TCD1
Ã9,1 = TCR1 XG1 /(TF1 R31YG1 bg1 )
Ã9,9 = −1/TF1
Ã9,10 = 1/TF1 + TCR1 /(TF1YG1 )
Ã10,1 = −XG1 /(R31YG1 bg1 )
Ã10,10 = −1/YG1
Ã10,11 = 1/YG1 − XG1 cg1 /(YG1 bg1 )
Ã11,1 = −1/(R31 bg1 )
Ã11,11 = −cg1 /bg1
Ã12,12 = −1/TPS2
Ã12,13 = KPS2 /TPS2
Ã12,16 = KPS2 /TPS2
Ã12,19 = KPS2 /TPS2
Ã12,23 = −a12 KPS2 /TPS2
TH-1171_07610212
128
Appendix A
Ã13,13 = −1/TT 2
Ã13,14 = 1/TT 2
Ã14,12 = −KR2 /(TSG2 R12 )
Ã14,14 = −1/TR2
Ã14,15 = 1/TR2 − KR2 /TSG2
Ã15,12 = −1/(TSG2 R12 )
Ã15,15 = −1/TSG2
Ã16,12 = 2TRS2 /(TRH2 TGH2 R22 )
Ã16,16 = −2/TW 2
Ã16,17 = 2/TW 2 + 2/TRH2
Ã16,18 = 2TRS2 /(TRH2 TGH2 ) − 2/TRH2
Ã17,12 = −TRS2 /(TRH2 TGH2 R22 )
Ã17,17 = −1/TRH2
Ã17,18 = 1/TRH2 − TRS2 /(TRH2 TGH2 )
Ã18,12 = −1/(TGH2 R22 )
Ã18,18 = −1/TGH2
Ã19,19 = −1/TCD2
Ã19,20 = 1/TCD2
Ã20,12 = TCR2 XG2 /(TF2 R32YG2 bg2 )
Ã20,20 = −1/TF2
Ã20,21 = 1/TF2 + TCR2 /(TF2YG2 )
Ã21,12 = −XG2 /(R32YG2 bg2 )
Ã21,21 = −1/YG2
Ã21,22 = 1/YG2 − XG2 cg2 /(YG2 bg2 )
Ã22,12 = −1/(R32 bg2 )
Ã22,22 = −cg2 /bg2
Ã23,1 = 2π T12
Ã23,12 = −2π T12
Ã24,1 = B1
Ã24,23 = 1
Ã25,12 = B2
Ã25,23 = a12
below:
B̃3,1 = KR1 ep f1 /TSG1
B̃4,1 = ep f1 /TSG1
B̃5,2 = −2TRS1 ep f2 /(TRH1 TGH1 )
B̃6,2 = TRS1 ep f2 /(TRH1 TGH1 )
B̃7,2 = ep f2 /TGH1
B̃9,3 = −XG1 TCR1 ep f3 /(TF1YG1 bg1 )
B̃10,3 = XG1 ep f3 /(YG1 bg1 )
B̃11,3 = ep f3 /bg1
TH-1171_07610212
129
Appendix A
B̃14,4 = KR2 ep f4 /TSG2

B̃15,4 = ep f4 /TSG2
B̃16,5 = −2TRS2 ep f5 /(TRH2 TGH2 )
B̃17,5 = TRS2 ep f5 /(TRH2 TGH2 )
B̃18,5 = ep f5 /TGH2
B̃20,6 = −XG2 TCR2 ep f6 /(TF2YG2 bg2 )
B̃21,6 = XG2 ep f6 /(YG2 bg2 )
B̃22,6 = ep f6 /bg2
below:
F̃1,1 = −KPS1 /TPS1
F̃12,2 = −KPS2 /TPS2
The matrix F̄ is of the dimension 25 × 4, where F̄i, j is the element of ith row and jth column of the matrix F̄. The non-zero elements are given
below:
F̄3,1 = cp f11 KR1 /TSG1
F̄3,2 = cp f12 KR1 /TSG1
F̄3,3 = cp f13 KR1 /TSG1
F̄3,4 = cp f14 KR1 /TSG1
F̄4,1 = cp f11 /TSG1
F̄4,2 = cp f12 /TSG1
F̄4,3 = cp f13 /TSG1
F̄4,4 = cp f14 /TSG1
F̄5,1 = −2TRS1 cp f21 /(TGH1 TRH1 )
F̄5,2 = −2TRS1 cp f22 /(TGH1 TRH1 )
F̄5,3 = −2TRS1 cp f23 /(TGH1 TRH1 )
F̄5,4 = −2TRS1 cp f24 /(TGH1 TRH1 )
F̄6,1 = cp f21 TRS1 /(TGH1 TRH1 )
F̄6,2 = cp f22 TRS1 /(TGH1 TRH1 )
F̄6,3 = cp f23 TRS1 /(TGH1 TRH1 )
F̄6,4 = cp f24 TRS1 /(TGH1 TRH1 )
F̄7,1 = cp f21 /TGH1
F̄7,2 = cp f22 /TGH1
F̄7,3 = cp f23 /TGH1
F̄7,4 = cp f24 /TGH1
F̄9,1 = TCR1 cp f31 XG1 /(bg1YG1 TF1 )
F̄10,1 = cp f31 XG1 /(bg1YG1 )
F̄10,2 = cp f32 XG1 /(bg1YG1 )
F̄10,3 = cp f33 XG1 /(bg1YG1 )
F̄10,4 = cp f34 XG1 /(bg1YG1 )
F̄11,1 = cp f31 /bg1
TH-1171_07610212
130
Appendix A
F̄11,2 = cp f32 /bg1

F̄11,3 = cp f33 /bg1
F̄11,4 = cp f34 /bg1
F̄14,1 = cp f41 KR2 /TSG2
F̄14,2 = cp f42 KR2 /TSG2
F̄14,3 = cp f43 KR2 /TSG2
F̄14,4 = cp f44 KR2 /TSG2
F̄15,1 = cp f41 /TSG2
F̄15,2 = cp f42 /TSG2
F̄15,3 = cp f43 /TSG2
F̄15,4 = cp f44 /TSG2
F̄16,1 = −2TRS2 cp f51 /(TGH2 TRH2 )
F̄16,2 = −2TRS2 cp f52 /(TGH2 TRH2 )
F̄16,3 = −2TRS2 cp f53 /(TGH2 TRH2 )
F̄16,4 = −2TRS2 cp f54 /(TGH2 TRH2 )
F̄17,1 = TRS2 cp f51 /(TGH2 TRH2 )
F̄17,2 = TRS2 cp f52 /(TGH2 TRH2 )
F̄17,3 = TRS2 cp f53 /(TGH2 TRH2 )
F̄17,4 = TRS2 cp f54 /(TGH2 TRH2 )
F̄18,1 = cp f51 /TGH2
F̄18,2 = cp f52 /TGH2
F̄18,3 = cp f53 /TGH2
F̄18,4 = cp f54 /TGH2
F̄21,1 = cp f61 XG2 /(bg2YG2 )
F̄21,2 = cp f62 XG2 /(bg2YG2 )
F̄21,3 = cp f63 XG2 /(bg2YG2 )
F̄21,4 = cp f64 XG2 /(bg2YG2 )
F̄22,1 = cp f61 /bg2
F̄22,2 = cp f62 /bg2
F̄22,3 = cp f63 /bg2
F̄22,4 = cp f64 /bg2
F̄24,1 = (cp f41 + cp f51 + cp f61 )
F̄24,2 = (cp f42 + cp f52 + cp f62 )
F̄24,3 = −(cp f13 + cp f23 + cp f33 )
F̄24,4 = −(cp f14 + cp f24 + cp f34 )
F̄25,1 = a12 (cp f41 + cp f51 + cp f61 )
F̄25,2 = a12 (cp f42 + cp f52 + cp f62 )
F̄25,3 = −a12 (cp f13 + cp f23 + cp f33 )
F̄25,4 = −a12 (cp f14 + cp f24 + cp f34 )
TH-1171_07610212
131
Appendix A

 
 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
 
 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 
C̃ = 
 

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 
 
 
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
below:
Q̃1,1 = 1 + B21
Q̃1,23 = B1
Q̃12,12 = 1 + B22
Q̃12,23 = B2 a12
Q̃23,1 = B1
Q̃23,12 = B2 a12
Q̃23,23 = 1 + a212
Q̃24,24 = 1
Q̃25,25 = 1
TH-1171_07610212
132
R EFERENCES
[1] A. Yazdizadeh, M. H. Ramezani, and E. Hamedrahmat, “Decentralized load frequency

control using a new robust optimal MISO PID controller,” Int. J. Electrical Power and
Energy Systems, vol. 35, pp. 57 –65, 2012.
[2] D. P. Kothari and I. J. Nagrath, Modern Power System Analysis, 4th ed. McGraw Hill,
2011.
[3] D. P. Kothari and J. S. Dhillon, Power System Optimization, 2nd ed. New Delhi, India:
Prentice Hall, 2010.
[4] Ibraheem, P. Kumar, and D. P. Kothari, “Recent philosophies of automatic generation

control strategies in power systems,” IEEE Trans Power Syst, vol. 20, no. 1, pp. 346
–357, 2005.
[5] H. Bevrani and T. Hiyama, Intelligent Automatic Generation Control. New York: CRC
Press Taylor and Francis, 2011.
[6] H. Bevrani, Robust Power System Frequency Control. New York: Springer, 2009.
[7] O. Elgerd, Electric Energy System Theory: an introduction, 2nd ed. New York:
McGraw-Hill, 1983.
[8] O. Elgerd and C. Fosha, “Optimum megawatt frequency control of multi-area electric
energy systems,” IEEE Trans Power Appl. Syst., vol. 89, no. 4, pp. 556 –563, 1970.
[9] L. C. Saikia, J. Nanda, and S. Mishra, “Performance comparison of several classical
controllers in AGC for multi-area interconnected thermal system,” Electrical Power and
Energy Systems, vol. 33, pp. 394 –401, 2011.
TH-1171_07610212
133
References
[10] P. Kundur, Power System Stability and Control, fifth reprint ed. New Delhi India: Tata
McGraw Hill, 2008.
[11] B. Tyagi and S. C. Srivastava, “A LQG based load frequency controller in a competitive
electricity environment,” Int. J. Emerging Elect. Power Syst., vol. 2, 2005. [Online].
Available: http://www.bepress.com/ijeeps/vol2/iss2/art1044
[12] A. Khodabakhshian and R. Hooshmand, “A new PID controller design for automatic
generation control of hydro power systems,” Int. J. Electrical Power and Energy Systems,
vol. 32, pp. 375 –382, 2010.
[13] T. A. Muthana, F. H. Mohamed, and Z. Mohamed, “Decentralized load frequency con-
troller for a multi-area interconnected power system,” Int. J. Electrical Power and Energy
Systems, vol. 33, pp. 198 –209, 2011.
[14] K. T. H. Bevrani, Y. Mitani, “Robust AGC: Traditional structure versus restructured

scheme,” IEEJ Trans. on Power and Energy, vol. 124, no. 5, pp. 751 –761, 2004.
[15] J. Nanda, S. Mishra, and L. C. Saikia, “Maiden application of bacterial foraging-based

optimization technique in multi-area automatic generation control,” IEEE Trans. Power
Syst, vol. 24, no. 2, pp. 602 –609, 2009.
[16] N. Jaleeli, D. N. Ewart, and L. H. Fink, “Understanding automatic generation control,”
IEEE Trans. Power System, vol. 7, no. 3, pp. 1106 –1112, 1992.
[17] H. Bevrani, “Decentralized robust load-frequency control synthesis in restructured power

systems,” Ph.D. dissertation, Department of Electrical Engineering Graduate School of
Engineering Osaka University, 2004.
[18] P. Bhatt, S. P. Ghoshal, R. Roy, and S. Ghosal, “Load frequency control of interconnected
restructured power system along with DFIG and coordinated operation of TCPS-SMES,”
IEEE Conf proc.,PEDES, 2010.
TH-1171_07610212
134
References
[19] P. Bhatt, R. Roy, and S. P. Ghoshal, “Optimized multiarea AGC simulation in restructured
power systems,” Int. Journal of Electrical Power and Energy Systems, vol. 32, no. 4, pp.
311 –322, 2010.
[20] V. Donde, M. A. Pai, and I. A. Hiskens, “Simulation and optimization in an AGC system
after deregulation,” IEEE Trans. Power Syst., vol. 16, no. 3, pp. 481 –489, 2001.
[21] R. D. Christie and A. Bose, “Load frequency control issues in power system operation
after deregulation,” IEEE Trans. Power Syst., vol. 11, no. 3, pp. 1191 –1200, 1996.
[22] N. Bekhouche, “Automatic generation control before after deregulation,” 34th Southeast-
ern Symp.System Theory, pp. 321 –323, 2002.
[23] C. Concordia and L. K. Kirchmayer, “Tie line power and frequency control of electric
power systems,” Amer. Inst. Elect. Eng. Trans., vol. pt. II,72, pp. 562 –572, 1953.
[24] L. K. Kirchmayer, Economic Control of Interconnected Systems. New York: Wiley,

1959.
[25] N. Cohn, “Some aspects of tie-line bias control on interconnected power systems,” Amer.
Inst. Elect. Eng. Trans., vol. 75, pp. 1415 –1436, 1957.
[26] ——, “Considerations in the regulation of interconnected area,” IEEE Trans. Power Syst.,
vol. PAS-86, pp. 1527 – 1538, 1967.
[27] J. E. Van Ness, “Root loci of load frequency control systems,” IEEE Trans. Power App.
Syst., vol. PAS-82, no. 5, pp. 712 –726, 1963.
[28] G. Quazza, “Noninteracting controls of interconnected electric power systems,” IEEE

Trans. Power App. Syst., vol. PAS-85, no. 7, pp. 727 –741, 1966.
[29] N. Cohn, “Techniques for improving the control of bulk power transfers on intercon-
nected systems,” IEEE Trans. Power App. Syst., vol. PAS-90, no. 6, pp. 2409 –2419,
1971.
TH-1171_07610212
135
References
[30] H. G. Kwatny, K. C. Kalnitsky, and A. Bhatt, “An optimal tracking approach to load
frequency control,” IEEE Trans. Power App. Syst., vol. PAS-94,, pp. 1635 –1643, 1975.
[31] C. Concordia, L. K. Kirchmayer, and E. A. Szymanski, “Effect of speed governor dead-

band on tie-line power and frequency control performance,” Amer. Inst. Elect. Eng.
Trans., vol. 76, pp. 429 –435, 1957.
[32] H. Golpira, H. Bevrani, and H. Golpira, “Application of GA optimization for automatic

generation control design in an interconnected power system,” Energy Conversion and
Management, vol. 52, pp. 2247 –2255, 2011.
[33] S. C. Tripathy, T. S. Bhatti, C. S. Jha, O. P. Malik, and G. S. Hope, “Sampled data
automatic generation control analysis with reheat steam turbines and governor dead band
effects,” IEEE Trans Power Appl. Syst., vol. 103(5), pp. 1045 –1051, 1984.
[34] D. Das, J. Nanda, M. L. Kothari, and D. P. Kothari, “Automatic generation control of

hydrothermal system with new area control error considering generation rate constraint,”
Elect. Mach. Power Syst., vol. 18, no. 6, pp. 461 –471, 1990.
[35] “IEEE power engineering systems committee report, dynamic models for steam and hy-
dro turbines for power systems studies,” IEEE Trans. Power App. Syst, vol. PAS-92,
1973.
[36] “IEEE power engineering systems committee report, hydraulic turbine and turbine con-
trol models for system dynamics,” IEEE Trans. Power. Syst., vol. PWRS-7, 1992.
[37] “Working group on prime mover and energy supply models for system dynamic per-
formance studies, dynamic models for combined cycle plants in power system studies,”
IEEE Trans. Power Syst., vol. 9, no. 3, pp. 1698 –1708, 1994.
[38] H. L. M and B. G. R, “Utility experience with gas turbine testing and modeling,” Pro-
ceedings of IEEE Power Engineering Society Winter Meeting, vol. 2, no. 2, pp. 671 –677,
2001.
TH-1171_07610212
136
References
[39] C. Fosha and O. Elgerd, “The megawatt frequency control problem: a new approach via
optimal control theory,” IEEE Trans Power Appl Syst, vol. 89, no. 4, pp. 563 – 577, 1970.
[40] E. C. Tacker, C. C. Lee, T. W. Reddoch, T. O. Tan, and P. M. Julich, “Optimal control of

interconnected electric energy systems: A new formulation,” Proc. IEEE, vol. 60, no. 10,
pp. 1239 – 1241, 1972.
[41] E. V. Bohn and S. M. Miniesy, “Optimum load frequency sample data control with ran-
domly varying system disturbances,” IEEE Trans. Power App. Syst., vol. PAS-91, no. 5,
pp. 1916 – 1923, 1972.
[42] K. Yamashita and T. Taniguchi, “Optimal observer design for load frequency control,”
Int. J. Elect. Power Energy Syst., vol. 8, no. 2, pp. 93 –100, 1986.
[43] A. Feliachi, “Load frequency control using reduced order models and local observers,”
Int. J. Energy Syst., vol. 7, no. 2, pp. 72 –75, 1987.
[44] A. Rubaai and V. Udo, “An adaptive control scheme for LFC of multiarea power systems.
part i: Identification and functional design, part-ii: Implementation and test results by
simulation,” Elect. Power Syst. Res, vol. 24, no. 3, pp. 183 –197, 1992.
[45] S. Velusami and K. Ramar, “Design of observer-based decentralized load-frequency con-

trollers for interconnected power systems,” Int. J. Power Energy Syst., vol. 17, no. 2, pp.
152 –160, 1997.
[46] Y. Hain, R. Kulessky, and G. Nudelman, “Identification-based power unit model for load-
frequency control purposes,” IEEE Trans. Power Syst., vol. 15, no. 4, pp. 1313 – 1321,
2000.
[47] V. R. Moorthi and R. P. Aggarawal, “Suboptimal and near optimal control of a load
frequency control system,,” Proc. Inst. Elect. Eng., vol. 119, pp. 1653 – 1660, Nov. 1972.
[48] S. S. Choi, H. K. Sim, and K. S. Tan, “Load frequency control via constant limited-state
feedback,” Elect. Power Syst. Res., vol. 4, no. 4, pp. 265 – 269, 1981.
TH-1171_07610212
137
References
[49] M. Aldeen and H. Trinh, “Load frequency control of interconnected power systems via
constrained feedback control schemes,” Int J Comput Elect Eng, vol. 20, no. 1, pp. 71
–88, 1994.
[50] G. Shirai, “Load frequency control using liapunovs second method: Bang-bang control
of speed changer position,” Proc. IEEE, vol. 67, no. 10, pp. 1458 – 1459, october 1979.
[51] A. Kumar, O. P.Malik, and G. S. Hope, “Discrete variable-structure controller for load
frequency control of multi-area interconnected power system .” Proc. Inst. Elect. Eng. C,
vol. 134, no. 2, pp. 116 –122, 1987.
[52] S. M. Miniesy and E. V. Bohn, “Two level control of interconnected power plants,” IEEE
Trans. Power App. Syst., vol. PAS-90, pp. 2742 –2748, June 1971.
[53] N. N. Bengiamin and W. C. Chan, “Multilevel load frequency control of interconnected

power system,” Proc. Inst. Elect. Eng., vol. 125, no. 6, pp. 521 –526, 1978.
[54] G. Shirai, “Load frequency control using generator and voltage controls via a new ap-
proach,” Proc. IEEE, vol. 66, no. 10, pp. 1293 – 1295, October 1978.
[55] N. Premakumaran, K. Parthasarathy, H. P. Khincha, and M. R. Chidambara, “Some as-

pects of multi-level load frequency control of a power system,” Proc. Inst. Elect. Eng. C,
vol. 129, no. 5, pp. 290 –294, november 1982.
[56] C. Ross, “A comprehensive direct digital load frequency controller,” Proc. IEEE, PICA
Conf., pp. 231 – 238, 1967.
[57] C. W. Ross and T. A. Green, “Dynamic performance evaluation of a computer controlled

electric power system,” IEEE Trans. Power App. Syst., vol. PAS-91, pp. 1156 – 1165,
1972.
[58] F. P. Demello, R. J. Mills, and W. F. BRells, “Automatic generation control, part iprocess
modeling,” IEEE Trans. Power App. Syst., vol. PAS-92, pp. 710 – 715, 1973.
TH-1171_07610212
138
References
[59] L. M. Smith, L. H. Fink, and R. P. Schulz, “Use of computer model of interconnected
power system to assess generation control strategies,” IEEE Trans. Power App. Syst.,
vol. 94, no. 5, 1975.
[60] C. W. Taylor and R. L. Cresap, “Real-time power system simulations for automatic gen-
eration control,” IEEE Trans. Power App. Syst., vol. PAS-95, pp. 375 – 384, 1976.
[61] A. Kumar, “Discrete load frequency control of interconnected power system,” Int. J.
Energy Syst., vol. 9, no. 2, pp. 73 – 77, 1989.
[62] M. L. Kothari, J. Nanda, D. P. Kothari, and D. Das, “Discrete mode automatic generation
control of a two area reheat thermal system with new area control error,” IEEE Trans.
Power App. Syst., vol. 4, no. 2, pp. 730 –738, May 1989.
[63] D. C. H. Prowse, “Improvements to a standard automatic generation control filter algo-

rithm,” IEEE Trans. Power Syst., vol. 8, no. 3, pp. 1204 – 1210, 1993.
[64] L. Hari, M. L. Kothari, and J. Nanda, “Optimum selection of speed regulation parame-
ters for automatic generation control in discrete mode considering generation rate con-
straints,” Proc. Inst. Elect. Eng. C, vol. 138, no. 5, pp. 401 – 406, 1991.
[65] Y. Wang, R. Zhou, and C. Wen, “Robust load-frequency controller design for power
systems,” Proc. Inst. Elect. Eng. C, vol. 140, no. 1, pp. 111 – 116, 1993.
[66] ——, “New robust adaptive load frequency control with system parameter uncertainties,”
Proc. Inst. Elect. Eng., vol. 141, no. 3, pp. 184 –190, May 1994.
[67] G. Ray and C. S. Rani, “Stabilizing decentralized robust controllers of interconnected

uncertain power systems based on the hessenberg form:simulated results,” Int. J. Syst.
Sci, vol. 32, no. 3, pp. 387 –399, 2001.
[68] T. C. Yang, Z. T. Ding, and H. Yu, “Decentralised power system load frequency control
beyond the limit of diagonal dominance,” Int. J. Elect. Power Energy Syst., vol. 24, pp.
173 –184, 2002.
TH-1171_07610212
139
References
[69] C. W. Ross, “Error adaptive control computer for interconnected power system,” IEEE
Trans. Power App. Syst., vol. PAS-85, p. 749, 1966.
[70] J. Kanniah, S. C. Tripathy, and O. P. Malik, “Microprocessor based adaptive load fre-
quency control,” Proc. Inst. Elect. Eng. C, vol. 131, no. 4, pp. 121 – 128, 1984.
[71] I. Vajk, M. Vajta, and L. Keviczky, “Adaptive load frequency control of hungarian power
system,” Automatica, vol. 21, no. 2, pp. 129 –137, 1985.
[72] C. T. Pan and C. M. Liaw, “An adaptive controller for power system and load frequency
control,” IEEE Trans. Power Syst., vol. 4, no. 1, pp. 122 –128, 1989.
[73] R. R. Shoults and J. A. J. Ibarra, “Multi-area adaptive LFC developed for a comprehen-
sive agc simulator,” IEEE Trans. Power App. Syst., vol. 8, no. 2, pp. 541 – 547, 1993.
[74] C. M. Liaw, “Design of a reduced-order adaptive LFC for an interconnected hydrother-

mal power system,” Int. J. Contr., vol. 60, no. 6, pp. 1051 – 1063, 1994.
[75] H. Shayeghi and H. A. Shayanfar, “Application of ANN technique for interconnected
power system load frequency control,” Int J Eng., vol. 16, no. 3, pp. 247 – 254, 2003.
[76] ——, “Application of ANN technique based on -synthesis to load frequency control of
interconnected power system,” Electr Power Energy Syst, vol. 28, pp. 503 – 511, 2006.
[77] S. Ramesh and A. Krishnan, “Fuzzy rule based load frequency control in a parallel AC-
DC interconnected power systems through hvdc link,” International Journal of Computer

Applications, vol. 1, no. 4, 2010.
[78] C. S. Chang and W. Fu, “Area load frequency control using fuzzy gain scheduling of PI
controllers,” Electr Power Syst Res, vol. 47, pp. 145 – 152, 1997.
[79] E. Cam and I. Kocaarslan, “Load frequency control in two area power system using fuzzy
logic controller,” J. Energy Conversion and Management, vol. 45, pp. 233 – 245, 2005.
TH-1171_07610212
140
References
[80] P. Bhatt, R. Roy, and S. Ghoshal, “GA/particle swarm intelligence based optimization
of two specific varieties of controller devices applied to two-area multi-units automatic

generation control,” Int. Journal of Electrical Power and Energy Syst., vol. 32, no. 4, pp.
299 – 310, May 2010.
[81] L. C. Saikia, N. Sinha, and J. Nanda, “Maiden application of bacterial foraging based
fuzzy idd controller in AGC of a multi-area hydrothermal system,” Electrical Power and
Energy Systems, vol. 45, pp. 98 – 106, 2013.
[82] M. Bettayeb, A. Quddus, and Randhawa, “Time-weighted optimal state and output feed-
back control of power systems,” Int. J. Electric Power Systems Research, vol. 52, pp. 77
– 86, 1999.
[83] R. Shahnazi and H. Khaloozadeh, “Output feedback control with disturbance rejection of
a class of nonlinear mimo systems,” Automatic Control and Computer Sciences, vol. 42,
no. 3, pp. 138 – 144, 2008.
[84] B. Tyagi and S. C. Srivastava, “A decentralized automatic generation control scheme for
competitive electricity markets,” IEEE Trans. on Power Systems, vol. 21, no. 1, pp. 312
–320, 2006.
[85] M. Shiroei, M. R. Toulabi, and A. M. Ranjbar, “Robust multivariable predictive based
load frequency control considering generation rate constraint,” Int. J. Electrical Power
and Energy Syst., vol. 46, pp. 405 – 413, 2013.
[86] N. Hasan, Ibraheem, P. Kumar, and Nizamuddin, “Sub-optimal automatic generation

control of interconnected power system using constrained feedback control strategy,”
Int. J. Electrical Power and Energy Syst., vol. 43, pp. 295 – 303, 2012.
[87] E. Rakhshani and J. Sadeh, “Reduced-order observer control for two-area LFC system
after deregulation,” Control Intell. Syst., vol. 38, no. 4, pp. 185 – 193, 2010.
TH-1171_07610212
141
References
[88] C. S. Chang, W. Fu, and F. Wen, “Load frequency control using genetic-algorithm based
fuzzy gain scheduling of PI controllers,” Electr Mach Power Syst, vol. 26, no. 1, pp. 39–
52, 1998.
[89] K. Challa and P. N. Rao, “Analysis and design of controller for two area thermal-hydro-
gas AGC system,” IEEE Conf. proceedings, PEDES, 2010.
[90] K. P. S. Parmar, S. Majhi, and D. P. Kothari, “Load frequency control of a realistic power
system with multi-source power generation,” International Journal of Electrical Power
and Energy Systems, vol. 42, pp. 426 –33, 2012.
[91] ——, “Automatic generation control of an interconnected hydrothermal power system,”
IEEE Conf. proceedings, INDICON, 2010.
[92] ——, “Multi-area load frequency control in a power system using optimal output feed-
back method,” IEEE Conf. proceedings, PEDES, 2010.
[93] ——, “Improvement of dynamic performance of LFC of the two area power system:
an analysis using MATLAB,” International Journal of Computer Applications, vol. 40,

no. 10, pp. 28 –32, February 2012.
[94] ——, “LFC of an interconnected power system with thyristor controlled phase shifter in
the tie line,” International Journal of Computer Applications, vol. 41, no. 9, pp. 27 –30,
2012.
[95] H. Shayeghi, H. Shayanfar, and A. Jalili, “Load frequency control strategies: A state-of-
the-art survey for the researcher,” Energy Conversion and Management, vol. 50, pp. 344
– 353, 2009.
[96] S. K. Aditya and D. Das, “Battery energy storage for load frequency control of an inter-
connected power system,” Electr Power Syst Res, vol. 58, no. 3, pp. 179 – 185, 2001.
TH-1171_07610212
142
References
[97] S. C. Tripathy, R. Balasubramanian, and P. S. C. Nair, “Effect of superconducting mag-
netic energy storage on automatic generation control considering governor dead-band

and boiler dynamics,” IEEE Trans Power Syst, vol. 7, no. 3, pp. 1266 – 1273, 1992.
[98] S. C. Tripathy, “Improved load frequency control with capacitive energy storage,” J.
Energy Conversion and Management, vol. 38, no. 6, pp. 551 – 562, 1997.
[99] M. Farahani and S. Ganjefar, “Solving LFC problem in an interconnected power system
using superconducting magnetic energy storage,” Physica C, vol. 487, pp. 60 – 66, 2013.
[100] I. Chidambaram and B. Paramasivam, “Optimized load-frequency simulation in restruc-

tured power system with redox flow batteries and interline power flow controller,” Int. J.
Electrical Power and Energy Syst., vol. 50, pp. 9 – 24, 2013.
[101] T. S. Bhatti and D. P. Kothari, “Variable structure load-frequency control of isolated

wind-diesel-microhydro hybrid power systems,” J. Institution of Engineers (India),
vol. 83, pp. 52 – 56, 2002.
[102] S. Ganapathy and S. Velusami, “Design of MOEA based decentralized load-frequency

controllers for interconnected power systems with AC-DC parallel tie-lines,” Interna-
tional Journal of Recent Trends in Engineering, vol. 2, no. 2, pp. 357 –361, 2009.
[103] P. Kumar and Ibraheem, “Dynamic performance evaluation of 2-area interconnected
power systems: a comparative study,” Electrical Engineering Division, J. Institution of

Engineers (India), vol. 78, pp. 199 –209, 1998.
[104] Ibraheem and P. Kumar, “Dynamic performance enhancement of hydropower systems

with asynchronous tie-lines,” J Electr Power Compon Syst, vol. 31, no. 7, pp. 605 – 626,
2003.
[105] K. Y. Lim, Y. Wang, and R. Zhou, “Decentralized robust load-frequency control in co-
ordination with frequency-controllable HVDC links,” Electr Power Energy Syst, vol. 19,
no. 7, pp. 423 – 431, 1997.
TH-1171_07610212
143
References
[106] K. Subbaramaiah, V. C. J. Mohan, and V. C. V. Reddy, “Comparison of performance of
SSSC and TCPS in automatic generation control of hydrothermal system under dereg-
ulated scenario,” International Journal of Electrical and Computer Engineering, vol. 1,
no. 1, pp. 21 –30, 2011.
[107] C. S. Rao, “Improvement of dynamic performance of AGC of hydrothermal system em-
ploying capacitive energy storage and TCPS,” Innovative Systems Design and Engineer-
ing, vol. 2, no. 6, pp. 63 –71, 2011.
[108] R. J. Abraham, D. Das, and A. Patra, “Damping oscillations in tie power and area fre-
quencies in a thermal power system with SMES-TCPS combination,” J. Electrical Sys-
tems, vol. 7, no. 1, pp. 71 –80, 2011.
[109] ——, “AGC of a hydrothermal system with thyristor controlled phase shifter in the tie-
line,” IEEE conf. proceedings, 2006.
[110] R. D. Christie and A. Bose, “Load frequency control in hybrid electric power markets,”
Proc. IEEE, Int. Conf. Contr. App, vol. 11, pp. 432 – 436, 1996.
[111] B. H. Bakken and O. S. Grande, “Automatic generation control in a deregulated power

system,” IEEE Trans. Power Syst., vol. 13, no. 4, pp. 1401 – 1406, 1998.
[112] J. Kumar, K. Ng, and G. Sheble, “AGC simulator for price-based operation part 1: A
model,” IEEE Trans.Power Syst., vol. 12, no. 2, pp. 527 –532, 1997.
[113] S. Chanana and A. Kumar, “A price based automatic generation control using unsched-
uled interchange price signals in indian electricity system,” International Journal of En-
gineering, Science and Technology, vol. 2, no. 2, pp. 23 – 30, 2010.
[114] S. Debbarma, L. C. Saikia, and N. Sinha, “AGC of a multi-area thermal system un-
der deregulated environment using a non-integer controller,” Electric Power Systems Re-
search, vol. 95, pp. 175 – 183, 2013.
TH-1171_07610212
144
References
[115] B. Tyagi and S. C. Srivastava, “Automatic generation control for multiarea system in a
deregulated electricity market,” Proc. Int. Conf. Bulk Power Transmission System Inte-
gration Developing Countries, Cigre Regional Meeting,, pp. VIII–18–VIII–29, 2001.
[116] K. S. S. Ramakrishna and T. S. Bhatti, “Automatic generation control of single area

power system with multi-source power generation,” Proc. IMechE: J. Power and Energy,
vol. 222 part A, pp. 1 –11, 2008.
[117] K. S. S. Ramakrishna, P. Sharma, and T. S. Bhatti, “Automatic generation control of inter-

connected power system with diverse sources of power generation,” Int. J. of Engineering
Science and Technology, vol. 2, no. 5, pp. 51 –65, 2010.
[118] T. E. Bechert and N. Chen, “Area automatic generation control by multi-pass dynamic
programming,” IEEE Trans. Power App. Syst., vol. PAS-96, pp. 1460 –1468, 1977.
[119] S. Mishra, G. Mallesham, and P. Sekhar, “Biogeography based optimal state feedback
controller for frequency regulation of a smart microgrid,” IEEE Transactions on Smart
Grid, vol. 4, no. 1, pp. 628 – 637, 2013.
[120] F. Llewis and V. L. Syrmos, Optimal Control. New Jersy , Englewood cliffs: Prentice
hall, 1995.
[121] Control Toolbox, MATLAB Software. MathWoks, Inc. MATLAB, vol. Version 7.13
(R2011b).
[122] Y. K. Singh and B. B. Chaudhuri, MATLAB Programming. New Delhi: Prentice hall,
2007.
[123] S. K. Sinha, “Automatic generation control in regulated and restructured power system,”
Ph.D. dissertation, IIT Roorkee,India, 2010.
[124] K. R. Sudha and R. V. Santhi, “Robust decentralized load frequency control of intercon-
nected power system with generation rate constraint using type-2 fuzzy approach,” Int.
J. Electrical Power and Energy Systems, vol. 33, pp. 699 –707, 2011.
TH-1171_07610212
145
References
[125] Y. L. Karanvas and D. P. Papadopoulos, “ AGC for autonomous power system using
combined intelligent techniques,” Electr Power Syst Res, vol. 62, no. 2, pp. 225 – 239,
2002.
[126] H. Shabani, B. Vahidi, and M. Ebrahimpour, “A robust PID controller based on imperi-
alist competitive algorithm for load-frequency control of power systems,” ISA Transac-
tions, vol. 52, pp. 88 – 95, 2013.
[127] C. S. Rao, S. S. Nagaraju, and P. Raju, “Improvement of dynamic performance of AGC

under open market scenario employing TCPS and AC-DC parallel tie line,” International
Journal of Recent Trends in Engineering, vol. 1, no. 1, 2009.
[128] H. D. Mathur and H. V. Manjunath, “Study of dynamic performance of thermal units

with asynchronous tie lines using fuzzy based controller,” J. Electrical Systems, vol. 3,
no. 3, pp. 124 –130, 2007.
TH-1171_07610212
146

TH-1171 07610212

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

TH-1171 07610212

Încărcat de

Drepturi de autor:

Formate disponibile

STATE SPACE BASED LOAD FREQUENCY CONTROL OF

MULTI-AREA POWER SYSTEMS

KRISHNA PAL SINGH PARMAR

in Partial Fulfilment of the Requirements

for the Degree of

Krishna Pal Singh Parmar

Department of Electronics and Electrical Engineering

Dated: Prof. Somanath Majhi

faculty members for their kind help during my academic studies.

(Krishna Pal Singh Parmar)

results are obtained.

List of Tables vii

List of Publications xiii

1.4 Significant contributions of the thesis . . . . . . . . . . . . . . . . . . . . . . . 10

2 Mathematical modeling and LFC of power systems 13

2.4 LFC modeling of the MAIPS . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.6.2 Selection of Q̃ and R̃ matrices . . . . . . . . . . . . . . . . . . . . . . 32

2.7 Important steps to solve LFC problem . . . . . . . . . . . . . . . . . . . . . . 32

2.8 State-space model of the power systems . . . . . . . . . . . . . . . . . . . . . 33

3 LFC of conventional power systems 35

3.2.2 Simulation results and discussion . . . . . . . . . . . . . . . . . . . . 38

3.4.1 Simulation results and discussion . . . . . . . . . . . . . . . . . . . . 46

4 LFC of power systems with multi-source power generation 51

4.2.2 Effect of generation rate constraint . . . . . . . . . . . . . . . . . . . . 55

4.3.1 Simulation results and discussion . . . . . . . . . . . . . . . . . . . . 63

4.3.5 Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

5 LFC of power systems considering AC-DC tie lines and TCPS 78

5.3.2 Simulation of the power system model considering TCPS . . . . . . . 89

6 LFC in restructured power system environment 98

6.3 Simulation and results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

7 Conclusions and Future work 115

A Matrices of power systems 119

2.2 Linear model of the primary LFC loop . . . . . . . . . . . . . . . . . . . . . . 18

interconnected power system with N-control areas . . . . . . . . . . . . . . . . 24

3.1 TAIPS with non-reheat turbines . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.2 Dynamic responses of TAIPS with non-reheat turbines . . . . . . . . . . . . . 41

4.4 Frequency deviation response to 1% SLP in the area considering GRC . . . . . 58

4.12 Dynamic responses considering the GRC . . . . . . . . . . . . . . . . . . . . 66

4.23 Model of hydro power plants operational in KHOZESTAN (Iran), [1] . . . . . 74

interconnected control areas w2 , second situation [1] . . . . . . . . . . . . . . 74

5.1 TAIPS with AC-DC parallel tie lines . . . . . . . . . . . . . . . . . . . . . . . 80

5.10 Frequency deviation response of control area-2 . . . . . . . . . . . . . . . . . 93

6.1 A TAIPS with MSPG in restructured power system environment . . . . . . . . 101

6.5 Generator power output response for case-2 . . . . . . . . . . . . . . . . . . . 110

3.1 Simulation parameters of TAIPS considering non-reheat turbines . . . . . . . . 39

3.3 Dynamic response comparison in terms of ST . . . . . . . . . . . . . . . . . . 40

3.7 Simulation parameters of hydro-thermal interconnected power system . . . . . 46

4.1 Simulation parameters of SAPS with MSPG . . . . . . . . . . . . . . . . . . . 55

4.2 Dynamic response comparison of a SAPS . . . . . . . . . . . . . . . . . . . . 56

5.7 Dynamic response comparison in terms of ST . . . . . . . . . . . . . . . . . . 96

ACE Area control error

AGC Automatic generation control

ISE Integral square error

MISO Multi-input single output

PI Proportional integral controller

ACEi Area control error

Prti Rated capacity of control area, MW

Yg (s) Governor dynamics

∆PDi Total incremental change in local load of control area, pu MW

∆Pvi Incremental change in governor valve position, pu MW

Hi Equivalent inertia constant of control area, MW-s/MVA

K̄ Full state feedback controller gain matrix